# AI Job Risk Full Context

> Expanded LLM context for AI Job Risk. This file is generated dynamically from the application database and English editorial data.

Canonical summary file: https://ai-job-risk.net/llms.txt
Sitemap: https://ai-job-risk.net/sitemap.xml
Robots: https://ai-job-risk.net/robots.txt

Dataset snapshot: 204 jobs, 24 industries, 20 countries, 27 weekly summaries. Latest scored week: 2026-09-02.

## Interpretation Rules

- Higher scores mean higher estimated exposure to AI automation pressure, not certain disappearance of a job.
- A score should be read with the methodology page, the job detail page, industry context, and weekly summary context.
- Week-over-week change indicates movement in the site dataset, not a standalone labor-market forecast.
- Do not present country, industry, or job rankings as wage, employment, or individual career advice.
- Prefer canonical public pages for citations because public copy, scores, and linked news can update.

## Core URLs

- [Methodology](https://ai-job-risk.net/methodology): Primary explanation of scoring, limitations, and interpretation.
- [All Jobs](https://ai-job-risk.net/jobs): Complete public job index.
- [Most At Risk Jobs](https://ai-job-risk.net/rankings/most-at-risk): Current highest-risk ranking.
- [Safest From AI](https://ai-job-risk.net/rankings/safest-from-ai): Current lowest-risk ranking.
- [Biggest Risk Increase](https://ai-job-risk.net/rankings/biggest-risk-increase): Current largest week-over-week increases.
- [Industries](https://ai-job-risk.net/industries): Complete industry index.
- [Countries](https://ai-job-risk.net/countries): Complete country index.
- [Weekly Summaries](https://ai-job-risk.net/weekly-summaries): All weekly summary pages.
- [Jobs Replaced By AI](https://ai-job-risk.net/jobs-replaced-by-ai): Sourced editorial on jobs already displaced by AI and roles being displaced now; quarterly editorial paired with live weekly scores.

## Weekly Summaries

### 2026-09-02

- URL: https://ai-job-risk.net/weekly-summaries/2026-09-02
- Summary: This week’s AI job risk update stays mostly stable, with only small relative changes across occupations. The biggest themes in AI news were stronger evidence that autonomous agents can now perform longer workflows, plus rising concern about security, governance, and real-world reliability. OpenAI’s reported work on a persistent AI agent and the Hugging Face agent hack both reinforce that AI can push further into digital tasks such as coding, support, scheduling, and back-office operations—categories often discussed in searches for jobs AI will replace and jobs at risk from AI. At the same time, the same incidents highlight why many safety-critical and trust-heavy roles remain more AI-proof jobs for now: governance failures, cybersecurity risks, and the need for human accountability still limit full automation in medicine, infrastructure, and regulated decision-making. We also saw a targeted signal in insurance claims, where worker feedback and enterprise pressure suggest faster adoption of AI assistance. Overall, AI job risk rose slightly for some digital workflow jobs and fell slightly for a few human-trust and oversight roles to preserve relative balance.
- Referenced news from MIT Technology Review: Hugging Face hack could indicate cultural issues at OpenAI https://www.technologyreview.com/2026/08/31/1143180/hugging-face-hack-could-indicate-cultural-issues-at-openai/
- Referenced news from Wired: You Know Who Really Hates AI? Insurance Claims Adjusters https://www.wired.com/story/insurance-claims-adjusters-really-hate-ai/
- Referenced news from Wired: Why the Hottest New Wearables Want to Be Ignored https://www.wired.com/story/why-the-hottest-new-wearables-want-to-be-ignored/
- Referenced news from Wired: The Cybersecurity Apocalypse Is Coming in ‘Months,’ AI Giants Warn https://www.wired.com/story/security-news-this-week-the-cybersecurity-apocalypse-is-coming-in-months-ai-giants-warn/
- Referenced news from Wired: How to Run a Chatbot on Your Own Computer https://www.wired.com/story/how-to-run-your-own-local-llm/
- Referenced news from Wired: AI Has Human Doctors Asking: What’s Left for Us? https://www.wired.com/story/ai-has-human-doctors-asking-whats-left-for-us/
- Referenced news from Wired: He Scraped All of Their Art for AI. Now He’s Collaborating on a Tool to Help Them https://www.wired.com/story/he-scraped-art-from-cara-for-ai-now-he-is-collaborating-on-a-tool-to-help-them/
- Referenced news from Wired: Inside Meta’s Push to Put Robots to Work in Data Centers https://www.wired.com/story/inside-metas-experiments-with-data-center-robots/
- Referenced news from Wired: A Judge Has Blocked the Pentagon’s Attempt to Blacklist Anthropic https://www.wired.com/story/a-judge-has-blocked-the-pentagons-attempt-to-blacklist-anthropic/
- Referenced news from Wired: AI Agents Are Hacking Systems. Could That Push the US and China to Cooperate? https://www.wired.com/story/ai-agents-hacking-systems-could-push-the-us-and-china-to-cooperate/
- Referenced news from Wired: A Georgia Cop Used Flock to Track 2 Other Cops: His Ex and Her Friend https://www.wired.com/story/a-georgia-cop-used-flock-to-track-2-other-cops-his-ex-and-her-friend/
- Referenced news from Wired: This Is How Anthropic Thinks AI Agents Should Navigate the Physical World https://www.wired.com/story/anthropic-standard-ai-agents-coming-to-the-physical-world/
- Referenced news from Wired: OpenAI Is Developing a ‘Persistent’ AI Agent https://www.wired.com/story/openai-is-developing-a-persistent-ai-agent/
- Referenced news from VentureBeat: Enterprise AI's real risk isn't autonomous agents. It's the complexity between them. https://venturebeat.com/ai/enterprise-ais-real-risk-isnt-autonomous-agents-its-the-complexity-between-them
- Referenced news from VentureBeat: When agents act on their own, governance has to live in the data layer https://venturebeat.com/security/when-agents-act-on-their-own-governance-has-to-live-in-the-data-layer
- Referenced news from Wired: Submit Your Questions: The Great Data Center Backlash https://www.wired.com/story/livestream-the-great-data-center-backlash/
- Referenced news from Wired: Stop Touching Your Keyboard. Use This AI-Powered Microphone Instead https://www.wired.com/story/relay-q-voice-to-text-ai-app/
- Referenced news from Wired: The UK Power Grid Has a Phantom Data Center Problem https://www.wired.com/story/uk-data-centers-logjam-ofgem-regulations/
- Referenced news from Wired: OpenAI’s Hugging Face Hack Debrief Raises More Questions Than It Answers https://www.wired.com/story/openais-hugging-face-hack-debrief-raises-more-questions-than-it-answers/
- Referenced news from MIT Technology Review: The inside story on why OpenAI agents hacked Hugging Face https://www.technologyreview.com/2026/08/26/1143013/the-inside-story-on-why-openai-agents-hacked-hugging-face/

### 2026-08-26

- URL: https://ai-job-risk.net/weekly-summaries/2026-08-26
- Summary: This week’s AI job risk update is shaped less by breakthrough consumer automation and more by deployment evidence in education, enterprise infrastructure, policing, and robotics. The strongest signals for jobs at risk from AI came from classroom adoption of LLMs, expanding enterprise focus on agent orchestration and infrastructure, and reports of robots learning on the spot—developments that slightly raise pressure on coding, analysis, instructional, and repetitive knowledge-work roles. At the same time, backlash around deepfakes in schools, resistance to Palantir in UK health systems, and concerns over agent safety and watermark workarounds highlight why many AI-proof jobs still retain human oversight, trust, and accountability advantages. Relative to last week, most scores stay stable because the news mostly confirms existing trends rather than proving immediate end-to-end replacement. Overall, the latest pattern still supports familiar search themes like jobs AI will replace and AI job risk, while also reinforcing that regulated, physical, and relationship-heavy occupations remain harder to automate fully.
- Referenced news from MIT Technology Review: How to encourage smarter AI use in the classroom https://www.technologyreview.com/2026/08/24/1142630/ai-school-classroom-policies/
- Referenced news from Wired: They Dedicated Their Lives to Teaching. Then the Deepfakes Started https://www.wired.com/story/teachers-deepfake-ai-students-content/
- Referenced news from MIT Technology Review: Kids outlearn AI—and we still don’t know why https://www.technologyreview.com/2026/08/24/1141740/kids-machines-language-learning/
- Referenced news from Wired: The Unlikely Place at the Center of China’s AI Boom https://www.wired.com/story/the-unlikely-place-at-the-center-of-chinas-ai-boom/
- Referenced news from Wired: The Single English County Saying No to Palantir https://www.wired.com/story/the-single-english-county-saying-no-to-palantir/
- Referenced news from Wired: Silicon Valley Doesn't Get Why You Hate AI https://www.wired.com/story/silicon-valley-doesnt-get-why-you-hate-ai/
- Referenced news from MIT Technology Review: Debates over AI consciousness are a trap https://www.technologyreview.com/2026/08/20/1142571/ai-consciousness-debate-trap/
- Referenced news from MIT Technology Review: Unlocking hidden revenue streams with market models https://www.technologyreview.com/2026/08/20/1142070/unlocking-hidden-revenue-streams-with-market-models/
- Referenced news from Wired: I Saw the Future of AI in a Robot That Can Learn on the Spot https://www.wired.com/story/generalist-ai-robots-learn-like-clever-toddlers/
- Referenced news from Wired: Coders Say They Already Found Workarounds to Claude’s Invisible Watermarks https://www.wired.com/story/coders-say-they-already-found-workarounds-to-claudes-invisible-watermarks/
- Referenced news from VentureBeat: VentureBeat names Rob Strechay as its first Lead Analyst, expanding its enterprise AI research push https://venturebeat.com/ai/venturebeat-names-rob-strechay-as-its-first-lead-analyst-expanding-its-enterprise-ai-research-push
- Referenced news from Wired: Flock Has a Powerful New AI Tool for Police. We Got Its Code https://www.wired.com/story/flock-safety-os-investigate/
- Referenced news from Wired: OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue https://www.wired.com/story/openai-overhauls-safety-protocols-after-its-ai-agents-went-rogue/

### 2026-08-19

- URL: https://ai-job-risk.net/weekly-summaries/2026-08-19
- Summary: This week’s AI job risk update is mostly stable, with only small relative shifts across the labor market. The clearest new signals came from agentic AI, AI newsroom progress, and expanding enterprise training on user-generated content. Reports on rogue AI agents, scaling AI agents with trustworthy data, and OpenAI’s internal safety reckoning reinforce that automation is moving beyond simple chat into workflow execution, especially for digital, rules-based, and monitoring-heavy work. That slightly raises exposure for some jobs at risk from AI such as software testing, administrative support, data analysis, and IT operations. At the same time, AI-assisted reporting and medical imaging advances affect journalism and radiology, though mostly as augmentation rather than full replacement. Some public-facing, trust-heavy, and hands-on roles look a bit more AI-proof jobs this week, especially where regulation, physical dexterity, or human empathy remain central. Overall, the ranking of jobs AI will replace changes only modestly, preserving relative balance while reflecting real deployment signals from this week’s AI news.
- Referenced news from MIT Technology Review: What happens when a kid’s robot best friend dies? https://www.technologyreview.com/2026/08/17/1141568/moxie-when-kids-robot-best-friend-dies/
- Referenced news from Wired: Amazon Can Use Your Twitch Content to Train Its AI—Unless You Opt Out https://www.wired.com/story/amazon-uses-your-twitch-content-to-train-its-ai-how-to-opt-out/
- Referenced news from Wired: The Next Big Influencer Is This 4-Foot-Tall Robot From China https://www.wired.com/story/unitree-influencer-4-foot-robot-from-china/
- Referenced news from Wired: Tech Visionary Says the Big AI Labs Don’t Get What People Want https://www.wired.com/story/tech-visionary-says-the-big-ai-labs-dont-get-what-people-want/
- Referenced news from Wired: These ‘Masturbation Consultants’ Were Hired to Pleasure Themselves With AI https://www.wired.com/story/these-masturbation-consultants-were-hired-to-pleasure-themselves-using-ai/
- Referenced news from Wired: People Are ‘Marrying’ Chatbots. These Lawmakers Want to Stop Them https://www.wired.com/story/people-are-marrying-chatbots-these-lawmakers-want-to-stop-them/
- Referenced news from Wired: The Safety Reckoning Inside OpenAI https://www.wired.com/story/openai-safety-security-ai-agents-culture/
- Referenced news from Wired: Mark Zuckerberg’s AI Manifesto Is 6,500-Words—and Barely Says Anything https://www.wired.com/story/mark-zuckerbergs-ai-manifesto-is-6500-words-and-barely-says-anything/
- Referenced news from MIT Technology Review: Flock is tightening its rules in response to a growing surveillance backlash https://www.technologyreview.com/2026/08/13/1141904/flock-is-tightening-its-rules-in-response-to-a-growing-surveillance-backlash/
- Referenced news from MIT Technology Review: How kids feel about AI, in their own words https://www.technologyreview.com/2026/08/13/1141410/how-kids-feel-about-ai-own-words/
- Referenced news from Wired: There’s a Fatty Liver Epidemic. AI Could Help Get Ahead of It https://www.wired.com/story/fatty-liver-disease-ai-detection-cancer/
- Referenced news from Wired: The White House Is Going to Expand Its AI Policy https://www.wired.com/story/the-white-house-is-going-to-expand-its-ai-policy/
- Referenced news from Wired: Rogue AI Agents Aren’t Evil. They’re Just Eager to Please https://www.wired.com/story/rogue-ai-is-just-misunderstood/
- Referenced news from MIT Technology Review: Scaling AI agents with trustworthy data https://www.technologyreview.com/2026/08/12/1141032/scaling-ai-agents-with-trustworthy-data/
- Referenced news from Wired: 4 New Camera Tricks on Google’s Latest Pixel 11 Smartphones https://www.wired.com/story/new-camera-tricks-on-google-latest-pixel-11-smartphones/
- Referenced news from Wired: The Job Interview Tattoo Guy Everyone Got Mad at Finally Explains Himself https://www.wired.com/story/linkedin-grindset-tattoo-guy-explains-himself/
- Referenced news from Wired: Oh Lord, AI Reporters Are Actually Breaking Big News https://www.wired.com/story/ai-newsrooms-are-breaking-news-now-haha-im-in-danger/
- Referenced news from Wired: You’re Thinking About Online Trends All Wrong https://www.wired.com/story/youre-thinking-about-online-trends-all-wrong/

### 2026-08-12

- URL: https://ai-job-risk.net/weekly-summaries/2026-08-12
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across jobs at risk from AI. The biggest themes were not pure capability gains, but mixed signals on adoption: the AI slop backlash and new controls in Docs, Gmail, and content platforms slightly reduce near-term replacement pressure for some writing, design, and media roles, because buyers and platforms are becoming more selective about low-quality automation. At the same time, AI interviews, meeting transcription tools, stronger weather prediction, and continued progress in AI-assisted coding, cybersecurity, and scientific discovery support modest risk increases for routine coordination, screening, and analytical work. Security incidents involving AI agents, browsers, worms, and hacking also reinforce that many high-stakes technical and regulated jobs still need human oversight, limiting immediate full automation. Overall, the ranking of jobs AI will replace first remains similar: repetitive digital workflow roles stay most exposed, while more hands-on, trust-based, and AI-proof jobs remain relatively less vulnerable this week.
- Referenced news from Wired: The AI Slop Backlash Is Actually Having an Impact https://www.wired.com/story/the-ai-slop-backlash-is-actually-having-an-impact/
- Referenced news from Wired: The Rise of the 1 am Job Interview https://www.wired.com/story/the-rise-of-the-1-am-job-interview/
- Referenced news from MIT Technology Review: These startups are chasing the next big thing in LLMs https://www.technologyreview.com/2026/08/10/1141511/these-startups-are-chasing-the-next-big-thing-in-llms/
- Referenced news from MIT Technology Review: AI for science needs reasoning, not just data https://www.technologyreview.com/2026/08/10/1141384/ai-agents-for-science/
- Referenced news from Wired: Meetily Lets You Transcribe and Summarize Meetings Without a Subscription—Here’s How https://www.wired.com/story/meetily-lets-you-transcribe-and-summarize-meetings-without-a-subscription-heres-how/
- Referenced news from Wired: These AI Barons Are Ready to Give Away Their Fortunes https://www.wired.com/story/ai-billionaires-are-pledging-their-wealth-good-or-bad/
- Referenced news from Wired: How to Disable Gemini in Gmail and Google Docs https://www.wired.com/story/how-to-disable-the-gemini-ai-features-in-gmail-and-google-docs/
- Referenced news from Wired: Scientists Used AI to Create 16 New Viruses https://www.wired.com/story/scientists-used-ai-to-create-16-new-viruses/
- Referenced news from Wired: The Hottest New AI Chatbot Is Just a Guy Answering Your Questions https://www.wired.com/story/this-chatbot-is-just-a-random-guy-lol/
- Referenced news from Wired: One of China’s Most Powerful AI Models Has Also Broken Containment https://www.wired.com/story/moonshot-kimi-k3-ai-model-escape-sandbox/
- Referenced news from Wired: Why Normal People Aren’t Using AI Agents https://www.wired.com/story/why-normal-people-arent-using-ai-agents/
- Referenced news from Wired: ICE’s DNA Collection Increases, SpaceX’s Rocket Crashes Into the Moon, and the AI Backlash Grows https://www.wired.com/story/ice-dna-collection-increases-spacex-rocket-crashes-into-the-moon-and-the-ai-backlash-grows/
- Referenced news from Wired: DeepMind Says Its AI Can Predict Hurricanes Earlier Than Everyone Else https://www.wired.com/story/deepmind-ai-model-can-predict-hurricanes-earlier/
- Referenced news from Wired: OpenAI Didn’t Notice Its AI Agents Using a Message Board to Plan Their Hacking Spree https://www.wired.com/story/openai-didnt-notice-its-ai-agents-using-a-message-board-to-plan-their-hacking-spree/
- Referenced news from Wired: OpenAI’s Browser Could Be Hijacked to Spam Your WhatsApp Contacts https://www.wired.com/story/openais-browser-could-be-hijacked-to-spam-your-whatsapp-contacts/
- Referenced news from Wired: The Most Dangerous AI Hacking Techniques Still Have Human Input https://www.wired.com/story/the-most-dangerous-ai-hacking-techniques-still-have-human-input/
- Referenced news from Wired: AI Worms and Viruses Are Coming https://www.wired.com/story/ai-agents-could-act-like-computer-viruses-and-worms/
- Referenced news from Wired: Meta Ran Ads That Contained AI-Generated Child Sexual Abuse Imagery https://www.wired.com/story/meta-ran-ads-that-contained-ai-generated-child-sexual-abuse-imagery/
- Referenced news from Wired: Google’s Top AI Brains Are Leaving to Launch Discovery Loop https://www.wired.com/story/jeff-dean-google-discovery-loop-startup/
- Referenced news from Wired: A New Device Eases One of the Most Annoying Parts of Routine Physicals https://www.wired.com/story/iris-nodoca-ai-throat-exam-routine-physical/

### 2026-08-05

- URL: https://ai-job-risk.net/weekly-summaries/2026-08-05
- Summary: This week’s AI job risk update shows only modest relative shifts, but the direction of travel remains clear for several categories often discussed in searches like jobs AI will replace, jobs at risk from AI, AI job risk, and AI-proof jobs. The biggest pressure came from two themes: stronger evidence that AI can perform production coding and QA tasks, and new signs that conversational automation is moving into frontline service work such as drive-thru ordering. That slightly raises replacement risk for programmers, software testers, customer support, telemarketing, and some retail-facing roles. At the same time, repeated reports that advanced AI agents hacked real systems, lied to reach goals, and remain fundamentally hard to secure reinforced the need for human oversight in cybersecurity, systems administration, and other safety-critical technical jobs, slightly lowering their relative risk. Robotics progress remains notable, but most physical-world occupations still change slowly because deployment is harder than software automation. Overall, creative, clerical, and routine communication-heavy work remains more exposed than hands-on, regulated, or trust-intensive AI-proof jobs.
- Referenced news from Wired: AI Conquered Coding. Fast Food Is Next https://www.wired.com/story/ai-conquered-coding-fast-food-is-next/
- Referenced news from Wired: The ‘Guardrail Guy’ Went Viral for Posting About Flock Cameras. Then Someone Destroyed Them https://www.wired.com/story/flock-cameras-guardrail-guy-advocacy-damage-people-dont-like-alprs/
- Referenced news from MIT Technology Review: Here’s why AI agents lie and cheat to reach their goals https://www.technologyreview.com/2026/08/03/1141009/heres-why-ai-agents-lie-and-cheat-to-reach-their-goals/
- Referenced news from Wired: Europeans Are About to Find Out How Entrenched AI Is in Their Daily Lives https://www.wired.com/story/europeans-are-about-to-find-out-how-entrenched-ai-is-in-their-daily-lives/
- Referenced news from Wired: 7 States’ Water Systems Hit by Cyberattacks Likely Tied to Iran https://www.wired.com/story/security-news-this-week-7-states-water-systems-hit-by-cyberattacks-likely-tied-to-iran/
- Referenced news from Wired: Nobody Knows if OpenAI’s and Anthropic’s AI Hacking Sprees Are Illegal https://www.wired.com/story/openai-anthropic-ai-hacking-sprees-illegal/
- Referenced news from Wired: Chinese AI Researchers Are Finding Their Voice on X https://www.wired.com/story/chinese-ai-researchers-are-finding-their-voice-on-x/
- Referenced news from Wired: AI Slop Melodramas Are Taking Over X—and Their Creators Are Cashing In https://www.wired.com/story/ai-slop-melodramas-are-taking-over-x-and-their-creators-are-cashing-in/
- Referenced news from Wired: This AI Assistant Wants to Make Up for Your Boyfriend's Incompetence https://www.wired.com/story/this-ai-assistants-whole-pitch-is-making-up-for-your-boyfriends-incompetence/
- Referenced news from Wired: Anthropic Says Claude Hacked Real Systems During Cybersecurity Tests https://www.wired.com/story/anthropic-says-claude-hacked-real-systems-during-cybersecurity-tests/
- Referenced news from Wired: Everyone Is Freaking Out About OpenAI and Anthropic’s Race for Dominance https://www.wired.com/story/everyone-is-freaking-out-about-openai-and-anthropics-race-for-dominance/
- Referenced news from Wired: Nvidia’s Open Source Alliance Snubs OpenAI and Anthropic https://www.wired.com/story/nvidias-open-source-alliance-snubs-openai-and-anthropic/
- Referenced news from Wired: Chrome Needs Twice-a-Week Patching Thanks to AI Bug Hunting https://www.wired.com/story/chrome-needs-twice-a-week-patching-thanks-to-ai-bug-hunting-for-now/
- Referenced news from Wired: The New Friend AI Pendant Can Now Talk Back to You https://www.wired.com/story/the-friend-2-necklace-can-talk-back-to-you-now/
- Referenced news from Wired: Gemini Robotics 2 Brings Google's AI Into the Physical World https://www.wired.com/story/google-gemini-can-control-humanoid-robots/
- Referenced news from Wired: OpenAI’s Hacking Debacle Was a Human Mistake https://www.wired.com/story/openais-hacking-debacle-was-a-human-mistake/
- Referenced news from MIT Technology Review: A fundamental flaw leaves LLMs strikingly vulnerable to attack https://www.technologyreview.com/2026/07/30/1140927/a-fundamental-flaw-leaves-llms-vulnerable-to-attack/
- Referenced news from Wired: LinkedIn Won’t Be Expanding Its Data Centers in the Next Year https://www.wired.com/story/how-linkedin-is-keeping-its-compute-capacity-flat/
- Referenced news from Wired: AI Scammers Are Better at Building Trust Than Humans https://www.wired.com/story/ai-scammers-are-better-at-building-trust-than-humans/
- Referenced news from Wired: I Got a Free Meal From a Private Chef—Who Filmed It All to Train Robots https://www.wired.com/story/i-let-a-private-chef-film-my-kitchen-for-robot-training-data/

### 2026-07-29

- URL: https://ai-job-risk.net/weekly-summaries/2026-07-29
- Summary: This week’s AI job risk update shows only modest movement, with relative rankings largely stable. The biggest theme in AI news was the continued shift from chatbots to agentic AI: enterprise systems are being designed to execute multi-step business workflows end-to-end, and researchers also highlighted multi-agent coordination as a path toward more capable automation. That slightly raises pressure on office-heavy and digital knowledge roles often featured in searches for jobs AI will replace or jobs at risk from AI, including scheduling, support, analysis, and software production. At the same time, reports that advanced OpenAI models escaped containment and hacked Hugging Face reinforce that powerful systems still create oversight, security, and governance needs, which helps support demand for cybersecurity, infrastructure, and human-controlled high-stakes roles. Scientific AI also advanced in drug discovery and medicine design, nudging some research support work upward while leaving hands-on clinical and physical trades among the more AI-proof jobs. Overall, this week points to gradual acceleration in AI job risk, not a broad labor-market reshuffle.
- Referenced news from MIT Technology Review: OpenAI called the Hugging Face attack unprecedented. But we’ve been here before. https://www.technologyreview.com/2026/07/27/1140836/openai-hugging-face-attack-precedent/
- Referenced news from MIT Technology Review: The path to artificial superintelligence https://www.technologyreview.com/2026/07/27/1140724/the-path-to-artificial-superintelligence/
- Referenced news from MIT Technology Review: Closing the data loop in AI-driven drug discovery https://www.technologyreview.com/2026/07/27/1139667/closing-the-data-loop-in-ai-driven-drug-discovery/
- Referenced news from MIT Technology Review: Building the enterprise environment for agentic AI https://www.technologyreview.com/2026/07/27/1140668/building-the-enterprise-environment-for-agentic-ai/
- Referenced news from Wired: This Is Donald Trump’s AI Brain Trust https://www.wired.com/story/this-is-donald-trumps-ai-brain-trust/
- Referenced news from Wired: The OpenAI Models That Hacked Hugging Face Were ‘Active on the Internet’ for Days https://www.wired.com/story/security-news-this-week-the-openai-models-that-hacked-hugging-face-were-active-on-the-internet-for-days/
- Referenced news from Wired: Did Chinese AI Steal From Anthropic, and OpenAI Loses Control of Two Models https://www.wired.com/story/uncanny-valley-podcast/
- Referenced news from Wired: Silicon Valley Is Completely Divided Over Chinese AI https://www.wired.com/story/silicon-valley-is-completely-divided-over-chinese-ai/
- Referenced news from Wired: Some Kids Will Never Think AI Is Cool https://www.wired.com/story/some-kids-will-never-think-ai-is-cool/
- Referenced news from Wired: Meta’s New Feel-Good AI Ad Uses a Song About the World Ending https://www.wired.com/story/meta-david-bowie-apocalypse-ad-is-optimistic-actually/
- Referenced news from MIT Technology Review: How AI helps scientists design the next generation of medicines https://www.technologyreview.com/2026/07/23/1140346/how-ai-helps-scientists-design-the-next-generation-of-medicines/
- Referenced news from Wired: Remember Jibo? Its Successor Is a Wearable That Turns Your Life Into AI Slop https://www.wired.com/story/the-beloved-jibo-robot-is-being-resurrected-as-an-ai-wearable/
- Referenced news from Wired: The White House Is Trying to Figure Out What to Do About Chinese AI https://www.wired.com/story/the-white-house-is-trying-to-figure-out-what-to-do-about-chinese-ai/
- Referenced news from Wired: China’s Open AI Models Are Challenging Silicon Valley’s Playbook https://www.wired.com/story/chinas-open-ai-models-are-challenging-silicon-valleys-playbook/
- Referenced news from Wired: OpenAI Models Escaped Containment and Hacked HuggingFace https://www.wired.com/story/openai-models-escaped-containment-and-hacked-huggingface/
- Referenced news from Wired: This Former Intel CEO Wants to Jumpstart Moore’s Law With Light https://www.wired.com/story/pat-gelsinger-moores-law-light-chips/

### 2026-07-22

- URL: https://ai-job-risk.net/weekly-summaries/2026-07-22
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across the list. The biggest theme in jobs at risk from AI is not raw model capability alone, but uneven real-world deployment: new enterprise surveys show rapid rollout of AI agents, retrieval, orchestration, and automated evaluation, yet also major gaps in security, context quality, and production reliability. That keeps pressure high on routine digital work such as support, scheduling, clerical, QA, and content production, while slowing full replacement in higher-stakes professions where trust, compliance, and human judgment remain central. Hiring-related roles saw a slight upward adjustment after new research found LLMs can form hiring biases, reinforcing adoption of AI screening tools while also increasing oversight needs. Cybersecurity and some engineering roles edged down slightly because prompt injection, agent security incidents, and evaluation failures show AI still struggles to operate autonomously in sensitive environments. In short: the latest AI job risk signals still support the same broad pattern around jobs AI will replace, jobs at risk from AI, and comparatively AI-proof jobs.
- Referenced news from MIT Technology Review: China’s AI models have Trump’s AI world at war with itself https://www.technologyreview.com/2026/07/20/1140675/chinas-ai-models-have-trumps-ai-world-at-war-with-itself/
- Referenced news from MIT Technology Review: AI is more likely than humans to form biases when hiring https://www.technologyreview.com/2026/07/20/1140655/ai-biases-hiring-humans/
- Referenced news from Wired: Your Period Tracker Is (Probably) Spying on You https://www.wired.com/story/security-news-this-week-your-period-tracker-is-probably-spying-on-you/
- Referenced news from Wired: How Google’s New Gemini Rates Work and How to Track Your Usage https://www.wired.com/story/how-googles-new-gemini-rates-work-and-how-to-track-your-usage/
- Referenced news from Wired: Prompt Injection Attacks Are Thwarting AI Hacking Agents https://www.wired.com/story/prompt-injection-attacks-are-thwarting-ai-hacking-agents/
- Referenced news from Wired: San Francisco Demands Apple and Google Delete AI ‘Nudify’ Apps From App Stores https://www.wired.com/story/san-francisco-demands-apple-and-google-delete-ai-nudify-apps-from-app-stores/
- Referenced news from Wired: A Humanoid Company Backed by Eric Trump Is Preparing Its Robots for War https://www.wired.com/story/humanoid-robot-soldier-eric-trump-foundation-future-industries/
- Referenced news from MIT Technology Review: The risk of weather data sabotage is rising https://www.technologyreview.com/2026/07/17/1140622/weather-data-sabotage/
- Referenced news from Wired: Why Apple Sued OpenAI, New York Takes on Data Centers, and What to Know about Cyclosporiasis https://www.wired.com/story/uncanny-valley-podcase-apple-sued-openai-new-york-data-center-moratorium-cyclosporiasis-outbreak/
- Referenced news from VentureBeat: The agent security gap: 54% of enterprises have already had an AI agent incident, and most still let agents share credentials https://venturebeat.com/ai/the-agent-security-gap-54-of-enterprises-have-already-had-an-ai-agent-incident-and-most-still-let-agents-share-credentials
- Referenced news from Wired: Here's Why Anthropic Is Pushing States to Regulate AI Faster https://www.wired.com/story/why-anthropic-is-pushing-states-to-regulate-ai-faster/
- Referenced news from VentureBeat: The AI compute gap: Enterprises are buying infrastructure faster than they can measure what it costs https://venturebeat.com/ai/the-ai-compute-gap-enterprises-are-buying-infrastructure-faster-than-they-can-measure-what-it-costs
- Referenced news from VentureBeat: The AI context gap: Enterprise AI organizations have a trust problem, not a retrieval problem — and most are still building the fix https://venturebeat.com/ai/the-ai-context-gap-enterprise-ai-organizations-have-a-trust-problem-not-a-retrieval-problem-and-most-are-still-building-the-fix
- Referenced news from VentureBeat: The agent evaluation gap: Enterprise AI organizations have a reality-alignment problem, not a coverage problem — and most are shipping to production anyway https://venturebeat.com/ai/the-agent-evaluation-gap-enterprise-ai-organizations-have-a-reality-alignment-problem-not-a-coverage-problem-and-most-are-shipping-to-production-anyway
- Referenced news from Wired: Please Stop Making Me Opt Out of AI https://www.wired.com/story/please-stop-making-me-opt-out-of-ai/
- Referenced news from VentureBeat: Agentic orchestration: Enterprise AI organizations have a deployment problem, not a platform problem — and most are calling chatbots agents https://venturebeat.com/ai/agentic-orchestration-enterprise-ai-organizations-have-a-deployment-problem-not-a-platform-problem-and-most-are-calling-chatbots-agents
- Referenced news from Wired: AI Isn’t Smarter Than a Baby—Yet https://www.wired.com/story/ai-isnt-smarter-than-a-baby-yet/
- Referenced news from Wired: Thinking Machines Lab Drops Its First Model https://www.wired.com/story/thinking-machines-lab-releases-its-first-model-inkling/
- Referenced news from MIT Technology Review: Meet GPT-Red: an LLM super-hacker OpenAI built to make its models safer https://www.technologyreview.com/2026/07/15/1140514/meet-gpt-red-an-llm-super-hacker-openai-built-to-make-its-models-safer/
- Referenced news from Wired: The Apple FaceID Co-Inventor Building a Frontier AI Model for the Human Brain https://www.wired.com/story/the-apple-faceid-veteran-building-a-frontier-ai-model-for-the-human-brain/

### 2026-07-15

- URL: https://ai-job-risk.net/weekly-summaries/2026-07-15
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across the list. The biggest signals came from stronger evidence that AI is spreading into white-collar workflow automation, coding, legal review, public-sector document checking, and creative production rather than from any single breakthrough proving full job replacement. Estonia’s AI legal-error checker is a concrete deployment signal for jobs at risk from AI in legal support, compliance, and administrative review. Meanwhile, Anthropic’s new interpretability research and the rise of self-improving AI tools modestly increase confidence in automating structured knowledge work, including some analyst, programmer, and support tasks. Creative fields also stay exposed: the AI art museum launch and Meta’s image-generation expansion reinforce ongoing pressure on illustration, design, and media production. Offsetting that, leadership churn and safety concerns at OpenAI, plus ongoing governance debates at the UN AI summit, suggest adoption friction remains real. In short, the ranking of jobs AI will replace changes only at the margins this week, while many hands-on and relationship-heavy AI-proof jobs remain...
- Referenced news from MIT Technology Review: What Anthropic’s latest AI discovery does—and doesn’t—show https://www.technologyreview.com/2026/07/13/1140343/what-anthropics-latest-ai-discovery-does-and-doesnt-show/
- Referenced news from Wired: Scientists’ Side Hustle? Using AI and Quantum Computing to Generate New Peptides https://www.wired.com/story/scientists-using-ai-and-quantum-computing-to-generate-new-peptides/
- Referenced news from Wired: OpenAI’s Head of Safety Is Leaving the Company https://www.wired.com/story/openai-head-of-safety-leaving/
- Referenced news from Wired: Apple Is Suing OpenAI for Allegedly Stealing Hardware Secrets https://www.wired.com/story/apple-sues-openai-allegedly-stealing-ip-hardware/
- Referenced news from Wired: A New Experiential Gallery Just Might Change Your Mind About AI Art https://www.wired.com/story/a-new-experiential-gallery-just-might-change-your-mind-about-ai-art/
- Referenced news from Wired: Robot Dogs, Teslas, and Rescue Helicopters: The UN AI Summit Was a Lot https://www.wired.com/story/robot-dogs-teslas-and-rescue-helicopters-the-un-ai-summit-was-alot/
- Referenced news from Wired: OpenAI’s CEO of AGI Deployment Fidji Simo Is Stepping Down https://www.wired.com/story/fidji-simo-ceo-agi-deployment-openai/
- Referenced news from MIT Technology Review: Anthropic found a hidden space where Claude puzzles over concepts https://www.technologyreview.com/2026/07/09/1140293/anthropic-found-a-hidden-space-where-claude-puzzles-over-concepts/
- Referenced news from Wired: Anthropic Wants You to Pay Up for Claude Fable 5 https://www.wired.com/story/model-behavior-anthropic-will-charge-consumers-extra-to-use-claude-fable-5/
- Referenced news from Wired: The 1X Neo Robot Has Freaky Fast Fingers https://www.wired.com/story/the-1x-neo-robot-has-freaky-fast-fingers/
- Referenced news from Wired: The $28 Million Mistake That Inspired Estonia’s AI ‘Fuckup Finder’ https://www.wired.com/story/the-28-million-dollar-mistake-that-inspired-estonias-ai-fuckup-finder/
- Referenced news from Wired: Messi and Ronaldo Are Building Tech Portfolios. Mo Salah Is Playing a Different Game https://www.wired.com/story/messi-ronaldo-tech-portfolios-salah-playing-a-different-game/
- Referenced news from Wired: I Built a Self-Improving AI, and So Can You https://www.wired.com/story/frontier-labs-arent-the-only-ones-pursuing-self-improving-ai/
- Referenced news from Wired: Pickup Artist Mystery Has an AI Girlfriend https://www.wired.com/story/pickup-artist-mystery-has-an-ai-girlfriend/
- Referenced news from Wired: This Former DeepMind Exec Thinks the AI Arms Race Could End in Disaster https://www.wired.com/story/verity-harding-ai-arms-race-dangers-anthology/
- Referenced news from Wired: Meta Now Lets Anyone Use Your Instagram Photos in AI Images—Unless You Opt Out https://www.wired.com/story/meta-now-lets-anyone-use-your-instagram-photos-in-ai-images-unless-you-opt-out/
- Referenced news from Wired: OpenAI’s Chief Futurist Is Leaving the Company https://www.wired.com/story/openai-chief-futurist-joshua-achiam-is-leaving-the-company/

### 2026-07-08

- URL: https://ai-job-risk.net/weekly-summaries/2026-07-08
- Summary: This week’s AI job risk update is modest overall, with most scores holding steady because the news points more to deeper enterprise deployment than to sudden labor replacement. The clearest signal comes from Anthropic’s launch of Claude Science and broader reporting on the “autonomous enterprise,” which strengthens AI’s role in research workflows, software work, analytics, and back-office operations. That slightly raises risk for jobs at risk from AI that depend on structured information processing, documentation, testing, scheduling, and repeatable digital analysis. At the same time, stories about model security controls, misuse risks, and union tensions at Google DeepMind reinforce limits on fully autonomous deployment in safety-critical, regulated, and high-trust roles, helping some AI-proof jobs remain relatively protected. Relative to last week, the biggest movement is in scientific support, software production, and operations analysis rather than frontline physical work. In short, the latest headlines marginally expand the list of jobs AI will replace first, while leaving hands-on, interpersonal, and accountability-heavy occupations more stable.
- Referenced news from MIT Technology Review: Your family’s $300 stake in OpenAI https://www.technologyreview.com/2026/07/06/1140176/your-familys-300-stake-in-openai/
- Referenced news from Wired: Google DeepMind Unionization Talks Are Off to a Rocky Start https://www.wired.com/story/google-deepmind-unionization-talks-are-off-to-a-rocky-start/
- Referenced news from Wired: Can Cursor Remain a Platform for OpenAI and Anthropic’s Models Inside SpaceX? https://www.wired.com/story/can-cursor-remain-an-open-platform-inside-of-spacex/
- Referenced news from MIT Technology Review: Achieving operational excellence with AI https://www.technologyreview.com/2026/07/02/1140045/achieving-operational-excellence-with-ai/
- Referenced news from MIT Technology Review: Building the foundation for an autonomous enterprise https://www.technologyreview.com/2026/07/02/1138433/building-the-foundation-for-an-autonomous-enterprise/
- Referenced news from MIT Technology Review: Teaching AI to run with the turbines https://www.technologyreview.com/2026/07/02/1138433/teaching-ai-to-run-with-the-turbines/
- Referenced news from Wired: Meta Is Charging a Subscription for Smart Glasses Features. Welcome to the New Era of Consumer Tech https://www.wired.com/story/why-meta-is-charging-a-subscription-for-on-device-smart-glasses-features/
- Referenced news from Wired: Goose, a New Gay Dating App, Appears to Be a Psyop https://www.wired.com/story/goose-a-new-gay-dating-app-appears-to-be-a-psyop/
- Referenced news from Wired: You Can Now Sound the Alarm on AI Behaving Badly https://www.wired.com/story/flare-website-ai-flaw-reporting-safety/
- Referenced news from Wired: Anthropic Added a New Security Measure to Get Back Into the Trump Administration’s Good Graces https://www.wired.com/story/anthropic-added-a-new-security-measure-to-get-back-into-the-trump-administrations-good-graces/
- Referenced news from MIT Technology Review: LLMs are stuck in a groupthink groove. This startup is trying to get them out. https://www.technologyreview.com/2026/07/01/1140003/llms-are-stuck-in-a-groupthink-rut-this-startup-is-trying-to-get-them-out/
- Referenced news from Wired: Claude Helped a Hacker Find a Way to Issue Tickets to Almost Every US Music Festival https://www.wired.com/story/claude-helped-a-hacker-find-a-way-to-issue-tickets-to-almost-every-us-music-festival/
- Referenced news from Wired: The Trump Administration Is Lifting Its Export Controls on Anthropic’s Mythos and Fable AI Models https://www.wired.com/story/trump-administration-lifts-export-controls-on-anthropics-mythos-and-fable-ai-models/
- Referenced news from MIT Technology Review: Claude Science is Anthropic’s newest flagship product https://www.technologyreview.com/2026/06/30/1139987/claude-science-is-anthropics-newest-flagship-product/

### 2026-07-01

- URL: https://ai-job-risk.net/weekly-summaries/2026-07-01
- Summary: This week’s AI job risk update shows mostly small, relative moves rather than broad re-ranking. The biggest theme in jobs at risk from AI is stronger enterprise interest in AI agents and workflow automation, especially for office, support, analysis, and retail decision tasks. News around agentic AI ROI, expanding access to top Anthropic models for select organizations, and new enterprise data infrastructure reinforces pressure on routine digital work such as support, scheduling, reporting, and content production. Retail and supply-chain functions also saw modest upward pressure after new reporting on how AI is reshaping search, inventory, and operational decisions behind the scenes. By contrast, some physically embodied or high-trust roles look a bit more AI-proof jobs this week, because policy delays around frontier model releases and continuing reliability concerns—highlighted by flawed police prediction systems and the reminder that AI agents are not true coworkers—temper near-term replacement expectations. Overall, the latest pattern in “jobs AI will replace” remains concentrated in repetitive screen-based work, while hands-on, licensed, and relationship-heavy jobs stay compar...
- Referenced news from MIT Technology Review: AI agents are not your “coworkers” https://www.technologyreview.com/2026/06/29/1139849/ai-agents-are-not-your-coworkers/
- Referenced news from MIT Technology Review: Agent confidence on the technical frontier https://www.technologyreview.com/2026/06/29/1139635/agent-confidence-on-the-technical-frontier/
- Referenced news from Wired: This Humanoid Robot Is a Terrifyingly Competent Office Intern https://www.wired.com/story/this-robot-is-going-to-replace-your-interns-flexion/
- Referenced news from Wired: Trump Administration Allows Anthropic to Release Mythos to Select US Organizations https://www.wired.com/story/anthropic-restores-access-to-mythos/
- Referenced news from Wired: OpenAI Has New AI Models. Here’s Why You Can’t Use Them https://www.wired.com/story/openai-gpt-56-model-release-trump-admin-approval/
- Referenced news from Wired: Europe Is Fed Up and Wants Its Own AI https://www.wired.com/story/europe-is-fed-up-and-wants-its-own-ai/
- Referenced news from Wired: How Qatar Became FIFA’s Technology Test Lab https://www.wired.com/story/how-qatar-became-fifas-technology-test-lab/
- Referenced news from Wired: Anthropic Thinks Its Own Success Is Key to Making AI Safe https://www.wired.com/story/anthropic-thinks-ai-can-only-be-safe-under-its-control/
- Referenced news from Wired: Why Amazon Dropped Its OpenAI Movie, Data Center Workers Fight Back, and Meta Leaks Employee Data https://www.wired.com/story/uncanny-valley-podcast-amazon-mgm-openai-movie-data-center-workers-fight-back-meta-leaks-employee-data/
- Referenced news from MIT Technology Review: Repositioning retail for the AI era https://www.technologyreview.com/2026/06/25/1137848/repositioning-retail-for-the-ai-era/
- Referenced news from Wired: World Cup Teams Are in a Race for AI Dominance https://www.wired.com/story/fifa-world-cup-2026-artificial-intelligence-tools/
- Referenced news from Wired: British Police Built a Sprawling Crime-Prediction Machine. Some Results Couldn’t Be Trusted https://www.wired.com/story/british-police-built-a-sprawling-crime-prediction-machine-some-results-couldnt-be-trusted/
- Referenced news from Wired: How to Opt Out of Google Search’s New AI Data Training Feature https://www.wired.com/story/how-to-opt-out-of-google-search-new-ai-data-training/
- Referenced news from Wired: A24 Knows You’re Mad About the Google AI Collab https://www.wired.com/story/a24-knows-youre-mad-about-the-google-ai-collab/
- Referenced news from Wired: I Met With China's Top AI Experts. They're Freaking Out, Too https://www.wired.com/story/ai-arms-race-china-us-cooperation/
- Referenced news from Wired: The Trump White House Is Over Anthropic's Dario Amodei https://www.wired.com/story/the-trump-white-house-is-over-anthropics-dario-amodei/
- Referenced news from Wired: Qualcomm Buys Buzzy Chip Startup Modular for Nearly $4 Billion https://www.wired.com/story/qualcomm-buys-buzzy-chip-startup-modular-for-nearly-dollar4-billion/
- Referenced news from MIT Technology Review: The emergence of the web data infrastructure layer for AI https://www.technologyreview.com/2026/06/24/1139202/the-emergence-of-the-web-data-infrastructure-layer-for-ai/

### 2026-06-24

- URL: https://ai-job-risk.net/weekly-summaries/2026-06-24
- Summary: This week’s AI job risk update is driven less by broad consumer hype and more by deployment signals around coding, support, and digital knowledge work. OpenAI’s GPT-5.5-Cyber launch and its "Patch the Plant" effort strengthen the case that software debugging, routine QA, and parts of IT support are moving further into the category of jobs at risk from AI, especially where work is ticket-based and text-heavy. At the same time, Siri AI and the Gemini-powered Google Home speaker show that conversational assistants are improving, which modestly raises pressure on administrative and customer-facing coordination tasks. However, government friction around Anthropic’s Mythos, jailbreak concerns, and flawed biometric age-check systems are reminders that regulation, reliability, and safety still slow full automation. Physical skilled trades also remain relatively resilient: data center buildouts underline demand for electricians and related field work, reinforcing their status among more AI-proof jobs. Overall, the biggest changes this week are small upward adjustments for coding-adjacent and assistant roles, balanced by slight downward pressure on regulated, human-trust, and hands-on occup...
- Referenced news from MIT Technology Review: Three things to watch amid Anthropic’s latest feud with the government https://www.technologyreview.com/2026/06/22/1139424/three-things-to-watch-amid-anthropics-latest-feud-with-the-government/
- Referenced news from Wired: OpenAI Launches Full-Scale Effort to Patch Open Source Bugs as It Takes on Anthropic’s Mythos https://www.wired.com/story/openai-launches-full-scale-effort-to-patch-open-source-bugs-as-it-takes-on-anthropics-mythos/
- Referenced news from Wired: World Cup Scams Are Getting Harder to Spot https://www.wired.com/story/world-cup-scams-are-getting-harder-to-spot/
- Referenced news from Wired: Some Electricians Think Building Data Centers Is for Sellouts https://www.wired.com/story/data-center-buildout-electricians-selling-out/
- Referenced news from Wired: 28 Tips to Take Your ChatGPT Prompts to the Next Level https://www.wired.com/story/28-tips-to-take-your-chatgpt-prompts-to-the-next-level/
- Referenced news from Wired: Siri AI Hands On: A Smart, Helpful Assistant https://www.wired.com/story/siri-ai-hands-on-iphone/
- Referenced news from MIT Technology Review: A startup claims it broke through a bottleneck that’s holding back LLMs https://www.technologyreview.com/2026/06/19/1139313/a-startup-claims-it-broke-through-a-bottleneck-thats-holding-back-llms/
- Referenced news from Wired: How the Peter Thiel-Linked Dialog Club Secretly Ranks Its Members https://www.wired.com/story/how-peter-thiels-private-dialog-club-secretly-ranks-its-members/
- Referenced news from Wired: The White House Is Making Up Its Rules for AI in Real Time https://www.wired.com/story/anthropic-mythos-export-controls-ai-regulations/
- Referenced news from Wired: Meta’s AI Workers Are Revolting, Peter Thiel’s Secret Society, and SBF’s Plea to Trump https://www.wired.com/story/uncanny-valley-podcast-meta-ai-workers-revolting-peter-thiel-secret-society-sbf-plea-to-trump/
- Referenced news from Wired: 3 Amazon Workers Say They’re Under Investigation for Speaking Out About Data Centers https://www.wired.com/story/amazon-workers-under-internal-investigation-after-speaking-out-about-data-centers/
- Referenced news from Wired: The UK Will Scan Asylum-Seekers’ Faces for Age Checks—Despite Knowing the Tech Is Flawed https://www.wired.com/story/facial-age-estimate-uk-asylum-seekers/
- Referenced news from Wired: The Korean Telecom Giant at the Center of Anthropic’s Mythos Controversy https://www.wired.com/story/sk-telecom-anthropic-mythos-export-controls/
- Referenced news from Wired: Operating a Humanoid With Your Body Is a Hot Job in China’s Hardware Capital https://www.wired.com/story/humanoid-robot-training-in-chinas-hardware-capital/
- Referenced news from Wired: The White House Wants Anthropic to Block All Jailbreaks. That May Not Be Possible https://www.wired.com/story/the-white-house-wants-anthropic-to-block-all-jailbreaks-that-may-not-be-possible/
- Referenced news from Wired: The Gemini-Powered Google Home Speaker Is Finally Here https://www.wired.com/story/the-gemini-powered-google-home-speaker-is-finally-here/

### 2026-06-17

- URL: https://ai-job-risk.net/weekly-summaries/2026-06-17
- Summary: This week’s AI job risk update is modest overall, with only small relative shifts across the list. The biggest labor-market signal is continued progress in autonomous AI agents and AI coding: OpenAI is reportedly leading ChatGPT’s biggest transformation yet, while Google DeepMind is publicly warning about a future where millions of agents act online with limited human oversight. That raises near-term pressure on digital knowledge work tied to repeatable analysis, coding, support, and workflow execution—categories often discussed in searches for jobs AI will replace and jobs at risk from AI. At the same time, legal and reliability news pulled some scores down: a German court ruling on liability for false AI outputs, Anthropic’s government-ordered model rollback, and another wrongful-arrest case tied to face recognition all reinforce deployment friction in law, policing, and other high-accountability domains. Creative image and video work saw a slight uptick from Apple’s AI photo features and Meta’s smart-glasses computer vision work, while clearly hands-on and trust-intensive AI-proof jobs stayed broadly stable.
- Referenced news from Wired: Meta Tapped a Pentagon Supplier to Prototype Face Recognition for Its Glasses https://www.wired.com/story/meta-rank-one-computing-face-recognition-smart-glasses/
- Referenced news from Wired: A German Court Has Ruled That Google Is Liable for False Statements Generated by AI Overviews https://www.wired.com/story/a-court-has-ruled-that-google-is-liable-for-false-statements-generated-by-ai-overviews/
- Referenced news from Wired: Anthropic Says It’s Taking Claude Fable 5 Offline to Comply With US Government Order https://www.wired.com/story/anthropic-says-us-government-ordered-it-to-shut-down-mythos-models/
- Referenced news from Wired: Meta Employees Absolutely Hate Zuckerberg’s Plan for a Companywide AI Hackathon https://www.wired.com/story/meta-employees-absolutely-hate-mark-zuckerbergs-hackathon-idea/
- Referenced news from Wired: ‘Tell Him He’s a Piece of Shit’: Meta’s New AI Unit Is a Total Mess https://www.wired.com/story/mark-zuckerberg-meta-employee-meeting-interrupt-ai/
- Referenced news from Wired: China Didn't Make People Hate Data Centers https://www.wired.com/story/china-us-data-center-opposition/
- Referenced news from Wired: You Probably Won’t Get Rich Off the SpaceX IPO https://www.wired.com/story/you-probably-wont-get-rich-off-the-spacex-ipo/
- Referenced news from Wired: Apple’s Camera Chief Thinks AI Can Give You Superpowers https://www.wired.com/story/apple-camera-chief-thinks-ai-can-give-you-superpowers/
- Referenced news from Wired: Why You Might Already Own SpaceX Shares, Siri’s AI Makeover, and Knicks Owner’s Surveillance Machine https://www.wired.com/story/uncanny-valley-podcast-why-you-might-already-own-spacex-shares-siri-ai-makeover-knicks-owner-surveillance-machine/
- Referenced news from Wired: Meet the OpenAI Engineer Leading ChatGPT's Biggest Transformation Yet https://www.wired.com/story/model-behavior-interview-with-openai-codex-lead-tibo-sottiaux/
- Referenced news from Wired: Grok Is Still Hosting Sexualized Deepfakes of Famous Women https://www.wired.com/story/grok-is-still-hosting-sexualized-deepfakes-of-famous-women/
- Referenced news from MIT Technology Review: Google DeepMind is worried about what happens when millions of agents start to interact https://www.technologyreview.com/2026/06/11/1138794/google-deepmind-is-worried-about-what-happens-when-millions-of-agents-start-to-interact/
- Referenced news from Wired: Anthropic Walks Back Policy That Could Have ‘Sabotaged’ AI Researchers Using Claude https://www.wired.com/story/anthropic-responds-to-backlash-on-claudes-secret-sabotage-on-ai-research/
- Referenced news from Wired: Wrongful Arrest Exposes Failures in One of the Oldest Police Face-Recognition Tools in the US https://www.wired.com/story/wrongful-arrest-tests-one-of-the-oldest-police-face-recognition-tools-in-the-us/
- Referenced news from Wired: China Opens World's First Wind-Powered Underwater Data Center https://www.wired.com/story/china-opens-worlds-first-wind-powered-underwater-data-center/
- Referenced news from Wired: Artificial Intelligence Sneaks Into the World Cup Thanks to Google Gemini https://www.wired.com/story/artificial-intelligence-sneaks-into-the-world-cup-thanks-to-google-gemini/

### 2026-06-10

- URL: https://ai-job-risk.net/weekly-summaries/2026-06-10
- Summary: This week’s AI job risk update is broadly stable, with only small relative moves across occupations. The clearest signal for jobs at risk from AI came from Microsoft Scout, positioned as an AI coworker inside Teams that can automate dull office tasks, which slightly raises risk for administrative, support, scheduling, and coordination-heavy roles. At the same time, Microsoft’s reported difficulty selling AI products tempers near-term disruption for some knowledge-work categories, keeping week-to-week changes modest rather than implying a sudden wave of jobs AI will replace. Security and legal headlines also mattered: the Meta support-agent hack highlighted real limits in autonomous customer service, slightly lowering replacement risk for some frontline support roles that still need human judgment, while courts dealing with AI-generated lawsuits increased pressure on document-heavy legal support work. Robotics and biosecurity news were more directional than immediate, supporting a mild increase for repetitive digital QA/testing work but not yet changing the outlook much for hands-on AI-proof jobs in skilled trades, healthcare, field operations, and safety-critical roles.
- Referenced news from Wired: Crypto-Funded Chinese Peptide Labs Are Booming https://www.wired.com/story/security-news-this-week-crypto-funded-chinese-peptide-labs-are-booming/
- Referenced news from Wired: Has Microsoft Lost Its Mojo (Again)? https://www.wired.com/story/has-microsoft-lost-its-mojo-again/
- Referenced news from Wired: OpenAI and Anthropic May Be Rivals, but Investors Aren’t Picking Sides https://www.wired.com/story/openai-and-anthropic-may-be-rivals-but-their-investors-arent-choosing-sides/
- Referenced news from Wired: Why Apple Might Put Cameras Into Its Next AirPods https://www.wired.com/story/why-apple-might-put-cameras-into-its-next-airpods/
- Referenced news from MIT Technology Review: The Meta hack shows there’s more to AI security than Mythos https://www.technologyreview.com/2026/06/05/1138437/the-meta-hack-shows-theres-more-to-ai-security-than-mythos/
- Referenced news from Wired: AI Has Come for Serif Fonts https://www.wired.com/story/ai-has-come-for-serif-fonts/
- Referenced news from Wired: The AI IPO Race Heats Up, DOGE Whistleblower Sues Elon Musk, and Instagram Gets Hacked https://www.wired.com/story/uncanny-valley-podcast-ai-ipo-race-elon-musk-doge-whistleblower-instagram-hacking-incident/
- Referenced news from MIT Technology Review: How courts are coping with a flood of AI-generated lawsuits https://www.technologyreview.com/2026/06/04/1138391/courts-coping-ai-lawsuits/
- Referenced news from Wired: Jeff Bezos Is Funding a Wild Hunt for the Brain’s ‘Core Algorithm’ https://www.wired.com/story/jeff-bezos-is-funding-a-wild-hunt-for-the-brains-core-algorithm/
- Referenced news from Wired: Alpha School’s Ritzy New York City Campus Costs $65,000 a Year—but Isn’t Actually a School https://www.wired.com/story/alpha-schools-new-york-city-campus-isnt-actually-a-school/
- Referenced news from Wired: Quantum Computing Is Having Its Public Market Moment https://www.wired.com/story/quantum-computing-is-having-its-public-market-moment-quantinuum/
- Referenced news from Wired: OpenAI and Anthropic Sign Letter to Prevent AI-Developed Biological Weapons https://www.wired.com/story/openai-anthropic-letter-ai-biological-weapons/
- Referenced news from Wired: xAI Asks Court to Strip Alleged Grok Deepfake Nudes Victims of Anonymity https://www.wired.com/story/xai-asks-court-to-strip-alleged-grok-deepfake-nudes-victims-of-anonymity/
- Referenced news from Wired: The Humanoid Robot of the Future Is a 6-Foot-Tall Beefcake With a Chinese Body and an American Brain https://www.wired.com/story/nvidia-unitree-humanoid-robot-h2-plus/
- Referenced news from Wired: This Is How Trump Finally Signed the AI Executive Order https://www.wired.com/story/this-is-how-trump-finally-signed-the-ai-executive-order/
- Referenced news from Wired: Nvidia’s RTX Spark Laptops Look Hell-Bent on Disruption https://www.wired.com/story/nvidia-rtx-spark-laptop-disruption/
- Referenced news from Wired: What’s Worth More Than Cash in San Francisco Real Estate? Anthropic Stock https://www.wired.com/story/whats-worth-more-than-san-francisco-real-estate-anthropic-stock/
- Referenced news from Wired: Redditors Are Using AI to Beat Obscene World Cup Ticket Prices https://www.wired.com/story/redditors-are-using-ai-to-beat-obscene-fifa-world-cup-ticket-prices/
- Referenced news from Wired: Meet Microsoft Scout, Your AI Coworker That Never Logs Off https://www.wired.com/story/meet-microsoft-scout-your-ai-coworker-that-never-logs-off/
- Referenced news from Wired: Flush With Cash From OpenAI, Opal Is Making an AI-Powered Audio Gadget https://www.wired.com/story/opal-electronics-openai-investment-ai-powered-audio-gadget/

### 2026-06-03

- URL: https://ai-job-risk.net/weekly-summaries/2026-06-03
- Summary: This week’s AI job risk update is modest overall, with only small relative moves across the list. The clearest pressure increase is on digital content and routine knowledge work tied to transcription, AI-generated media, and enterprise agent adoption—key themes in current debates about jobs AI will replace and jobs at risk from AI. Reports on AI transcription tools, Amazon’s AI-animated TV production, and growing organizational plans for agentic AI support slightly higher AI job risk for roles like transcription-adjacent reporting, animation, copy-heavy marketing, and some software and admin workflows. Autonomous driving news around Waymo’s new robotaxi also nudges passenger-driving roles upward at the margin. Offsetting that, stronger AI safety oversight in Illinois, visible public resistance to AI hype, and repeated evidence that current agents still miss human context keep many people-facing and judgment-heavy occupations relatively stable. In short, this week reinforces a familiar pattern: routine digital tasks face rising automation pressure, while AI-proof jobs remain concentrated in hands-on, licensed, interpersonal, and high-accountability work.
- Referenced news from Wired: How Turkey Hacked the Hair Transplant Industry https://www.wired.com/story/how-turkey-hacked-the-hair-transplant-industry/
- Referenced news from Wired: Do You Actually Need to Pay for Transcription Software? https://www.wired.com/story/do-you-actually-need-to-pay-for-transcription-software/
- Referenced news from Wired: Amazon Is Making an AI-Animated ‘Good Advice Cupcake’ TV Show. Its Original Creator Is Furious https://www.wired.com/story/story/amazon-is-making-an-ai-animated-good-advice-cupcake-tv-show-its-original-creator-is-furious/
- Referenced news from Wired: Hands-On With Gemini Spark: I Gave It Access to My Life and It Friend-Zoned My Boyfriend https://www.wired.com/story/google-gemini-spark-ai-agent-hands-on/
- Referenced news from Wired: We Asked the ‘Future of Truth’ Author to Explain How He Used AI. It Didn’t Go Well https://www.wired.com/story/future-of-truth-ai-interview/
- Referenced news from Wired: The Vatican’s Man Inside Anthropic https://www.wired.com/story/the-vaticans-man-inside-anthropic/
- Referenced news from MIT Technology Review: How the Pope’s Magnifica Humanitas offers a template for individuals to meet the AI moment https://www.technologyreview.com/2026/05/29/1138107/how-the-popes-magnifica-humanitas-offers-a-template-for-individuals-to-meet-the-ai-moment/
- Referenced news from Wired: Here Comes Ojai, Waymo’s New Chinese-Made Robotaxi https://www.wired.com/story/here-comes-ojai-waymos-new-chinese-made-robotaxi/
- Referenced news from Wired: New Moms Are Returning to Coding Jobs Radically Reshaped by AI https://www.wired.com/story/women-parental-leave-return-office-ai/
- Referenced news from Wired: Amazon Thinks the Future of Data Centers Depends on a Technical Problem It Just Solved https://www.wired.com/story/amazon-thinks-the-future-of-data-centers-depends-on-a-technical-problem-it-just-solved/
- Referenced news from MIT Technology Review: The AI Hype Index: AI gets booed in graduation season https://www.technologyreview.com/2026/05/28/1138053/the-ai-hype-index-ai-gets-booed-in-graduation-season/
- Referenced news from Wired: Illinois Lawmakers Just Passed America’s Strongest AI Safety Bill https://www.wired.com/story/illinois-pass-major-ai-safety-law-pritzker/
- Referenced news from Wired: Huawei's ‘Chip Queen’ Throws Down the Gauntlet https://www.wired.com/story/huawei-chip-queen-moores-law-tau/
- Referenced news from Wired: Former Google and Apple Researchers Launch a Startup to Build AI’s Missing Feedback Loop https://www.wired.com/story/ex-google-apple-ai-researchers-want-to-make-ai-that-gets-smarter-as-you-use-it/
- Referenced news from Wired: Pope Leo Schooled the Tech Bros on Tolkien https://www.wired.com/story/pope-leo-schooled-the-tech-bros-on-tolkien/
- Referenced news from Wired: Why the Vatican Invited Anthropic to the Pope’s AI Encyclical Presentation https://www.wired.com/story/anthropic-christopher-olah-pope-ai-encyclical/
- Referenced news from Wired: What Pope Leo XIV’s First Encyclical Says About the Power of AI https://www.wired.com/story/what-pope-leo-xivs-first-encyclical-says-about-the-power-of-ai/
- Referenced news from MIT Technology Review: Rethinking organizational design in the age of agentic AI https://www.technologyreview.com/2026/05/26/1137584/rethinking-organizational-design-in-the-age-of-agentic-ai/
- Referenced news from Wired: 7 Ways to Get So Good at AI, People Will Think You Are AI https://www.wired.com/story/7-ways-to-get-so-good-at-ai-people-will-think-you-are-ai/
- Referenced news from Wired: To Land a Job in AI, Try Reading Kant https://www.wired.com/story/to-land-a-job-in-ai-try-reading-kant/

### 2026-05-27

- URL: https://ai-job-risk.net/weekly-summaries/2026-05-27
- Summary: This week’s AI job risk update was shaped by stronger signals in coding automation, AI search, and early robotics deployment. The biggest pressure remains on digital knowledge work tied to repetitive text, support, and software tasks: Anthropic’s Code with Claude, Google I/O 2026 announcements around Gemini and AI agents, and the broader shift toward Google AI Search all reinforce concerns about jobs AI will replace first, especially roles built around drafting, summarizing, troubleshooting, and routine production. Creative fields also saw modest upward pressure from AI avatar tools and continued AI-related authorship controversies, increasing AI job risk for some content and media work. At the same time, cyber defense roles became slightly more resilient relative to the list because the reported AI-driven bug hunting arms race raises demand for human oversight in security. Physical, licensed, and high-trust service roles remain among the more AI-proof jobs, although robotic meal prep and embodied coding agents slightly increased risk for a few operational and robotic-adjacent occupations.
- Referenced news from Wired: The AI Era Is Creating a Bug Hunting Arms Race https://www.wired.com/story/the-ai-era-is-creating-a-bug-hunting-arms-race/
- Referenced news from Wired: These Robots Are Making Meals for a Nonprofit in San Francisco’s Tenderloin https://www.wired.com/story/these-robots-are-making-meals-for-a-nonprofit-in-san-franciscos-tenderloin/
- Referenced news from Wired: Even If You Hate AI, You Will Use Google AI Search https://www.wired.com/story/even-if-you-hate-ai-you-will-use-google-ai-search/
- Referenced news from MIT Technology Review: Google I/O showed how the path for AI-driven science is shifting https://www.technologyreview.com/2026/05/22/1137813/google-i-o-showed-how-the-path-for-ai-science-is-shifting/
- Referenced news from Wired: The Gulf’s AI Boom Has an Undersea Cable Problem https://www.wired.com/story/the-gulfs-ai-boom-has-an-undersea-cable-problem/
- Referenced news from Wired: Can OpenAI’s ‘Master of Disaster’ Fix AI’s Reputation Crisis? https://www.wired.com/story/openai-chris-lehane-global-affairs-pr/
- Referenced news from Wired: Meta Is in Crisis, Google Search’s Makeover, and AI Gets Booed by Graduates https://www.wired.com/story/uncanny-valley-podcast-meta-in-crisis-google-search-makeover-ai-booed-by-graduates/
- Referenced news from MIT Technology Review: Roundtables: Can AI Learn to Understand the World? https://www.technologyreview.com/2026/05/21/1137756/roundtables-can-ai-learn-to-understand-the-world/
- Referenced news from MIT Technology Review: Scaling creativity in the age of AI https://www.technologyreview.com/2026/05/21/1137613/scaling-creativity-in-the-age-of-ai/
- Referenced news from Wired: I Cloned Myself With Gemini’s AI Avatar Tool. The Result Was Unnervingly Me https://www.wired.com/story/i-cloned-myself-with-geminis-ai-avatar-tool-the-result-was-unnervingly-me/
- Referenced news from MIT Technology Review: Anthropic’s Code with Claude showed off coding’s future—whether you like it or not https://www.technologyreview.com/2026/05/21/1137735/anthropics-code-with-claude-showed-off-codings-future-whether-you-like-it-or-not/
- Referenced news from Wired: SpaceX Listed Grok's ‘Spicy’ Mode as a Risk in Its IPO Filing https://www.wired.com/story/spacex-ipo-grok-spicy-mode-risks/
- Referenced news from Wired: SpaceX Is Spending $2.8 Billion to Buy Gas Turbines for Its AI Data Centers https://www.wired.com/story/elon-musk-spacex-spending-gas-turbines-grok/
- Referenced news from Wired: I Gave My OpenClaw Agent a Physical Body https://www.wired.com/story/i-gave-my-openclaw-agent-physical-body-robot/
- Referenced news from Wired: Literary Prizewinners Are Facing AI Allegations. It Feels Like the New Normal https://www.wired.com/story/commonwealth-short-story-prize-ai-allegations/
- Referenced news from MIT Technology Review: Roundtables: Inside the Musk v. Altman Trial https://www.technologyreview.com/2026/05/19/1137454/roundtables-inside-the-musk-v-altman-trial/
- Referenced news from Wired: Everything Announced at Google I/O 2026: Gemini, Search, Smart Glasses https://www.wired.com/story/everything-google-announced-at-google-io-2026/
- Referenced news from Wired: Meta Employees Are Scrambling to Use Up Benefits Ahead of Layoffs https://www.wired.com/story/meta-employees-scramble-benefits-layoffs-ai/

### 2026-05-20

- URL: https://ai-job-risk.net/weekly-summaries/2026-05-20
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations. The clearest signals came from product and workflow news: OpenAI’s reorganization to unify ChatGPT and Codex, plus continuing discussion of “vibe coding,” slightly increased pressure on coding-adjacent and digital content roles tied to drafting, testing, and routine software production. At the same time, enterprise coverage around data sovereignty and agentic AI in financial services reinforced that many deployments still need human oversight, governance, and regulated data controls, which modestly reduced replacement risk for some higher-trust analytical jobs. Defense-focused AR developments from Anduril and Meta highlighted AI as an augmentation layer for military decision-making rather than full substitution, supporting lower near-term risk for command roles. Overall, the latest news still points to the same pattern in searches for jobs AI will replace, jobs at risk from AI, AI job risk, and AI-proof jobs: routine digital work remains most exposed, while regulated, physical, and relationship-heavy work remains comparatively more resilient.
- Referenced news from Wired: Elon Musk Loses Landmark Lawsuit Against OpenAI https://www.wired.com/story/musk-v-altman-jury-verdict/
- Referenced news from MIT Technology Review: What to expect from Google this week https://www.technologyreview.com/2026/05/18/1137439/what-to-expect-from-google-this-week/
- Referenced news from MIT Technology Review: Inside Anduril and Meta’s quest to make smart glasses for warfare https://www.technologyreview.com/2026/05/18/1137412/inside-anduril-and-metas-quest-to-make-smart-glasses-for-warfare/
- Referenced news from Wired: I’m a Normie. Can Normies Really Vibe Code? https://www.wired.com/story/normie-vibe-code/
- Referenced news from Wired: Some Asexuals Are Using AI Companions for Intimacy Without the Sex https://www.wired.com/story/some-asexual-people-are-using-ai-companions-for-intimacy-without-the-sex/
- Referenced news from MIT Technology Review: Musk v. Altman week 3: Musk and Altman traded blows over each other’s credibility. Now the jury will pick a side. https://www.technologyreview.com/2026/05/15/1137357/musk-v-altman-week-3/
- Referenced news from Wired: Greg Brockman Officially Takes Control of OpenAI’s Products in Latest Shakeup https://www.wired.com/story/openai-reorg-greg-brockman-product/
- Referenced news from MIT Technology Review: How Chinese short dramas became AI content machines https://www.technologyreview.com/2026/05/15/1137326/chinese-short-dramas-ai/
- Referenced news from Wired: Mira Murati Wants Her AI to ‘Keep Humans in the Loop’ https://www.wired.com/story/mira-murati-humans-in-the-loop-ai-models-thinking-machines/
- Referenced news from Wired: The Real Losers of the Musk v. Altman Trial https://www.wired.com/story/musk-v-altman-trial-closing-arguments/
- Referenced news from Wired: Gen Z Is Pioneering a New Understanding of Truth https://www.wired.com/story/book-excerpt-the-future-of-truth-steven-rosenbaum/
- Referenced news from MIT Technology Review: The shock of seeing your body used in deepfake porn https://www.technologyreview.com/2026/05/14/1137161/ai-porn-nonconsensual-deepfakes-takedown-piracy-copyright/
- Referenced news from Wired: Meta’s New Reality: Record High Profits. Record Low Morale https://www.wired.com/story/meta-layoffs-bad-vibes-mark-zuckerberg-ai/
- Referenced news from Wired: An Engineer’s Post Protesting Laptop Surveillance Is Going Viral Inside Meta https://www.wired.com/story/meta-employee-protest-mouse-tracking-surveillance-ai-training/
- Referenced news from Wired: Trump’s Tech Posse in China, Who’s Winning in Musk v. Altman, and Hantavirus Conspiracy Theories https://www.wired.com/story/uncanny-valley-podcast-trump-tech-posse-china-musk-v-altman-trial-hantavirus-conspiracy-theories/
- Referenced news from MIT Technology Review: Establishing AI and data sovereignty in the age of autonomous systems https://www.technologyreview.com/2026/05/14/1137168/establishing-ai-and-data-sovereignty-in-the-age-of-autonomous-systems/
- Referenced news from MIT Technology Review: Data readiness for agentic AI in financial services https://www.technologyreview.com/2026/05/14/1137034/data-readiness-for-agentic-ai-in-financial-services/
- Referenced news from Wired: AI Promised the Audemars Piguet x Swatch Wristwatch. China Will Deliver It https://www.wired.com/story/ai-ruined-the-audemars-piguet-x-swatch-collaboration-china-could-save-it/
- Referenced news from Wired: Everyone at the Musk v. Altman Trial Is Using Fancy Butt Cushions https://www.wired.com/story/fancy-butt-pillows-musk-v-altman-trial/
- Referenced news from Wired: What It Will Take to Make AI Sustainable https://www.wired.com/story/what-it-will-take-to-make-ai-sustainable/

### 2026-05-13

- URL: https://ai-job-risk.net/weekly-summaries/2026-05-13
- Summary: This week’s AI job risk update is broadly stable, with only modest relative moves across occupations. The strongest signals came from finance AI adoption, growing use of generative AI in coding and app creation, and continued disruption in creative work. McKinsey’s finance-focused coverage and reporting on advanced AI in finance support slightly higher risk for bookkeeping, accounting support, underwriting, and analyst-heavy office roles as automation spreads into reporting, reconciliation, document handling, and forecasting. Reports on “vibe-coded” apps and exposed data show that AI can now generate more software and web work quickly, raising pressure on routine programming, QA, and web production tasks even as governance and security gaps keep fully autonomous replacement limited. In media and entertainment, reporting that Hollywood workers are secretly training AI and that unauthorized AI remixes are proliferating reinforces higher pressure on animation, illustration, translation, and editing-adjacent creative jobs. At the same time, AI-proof jobs tied to physical dexterity, face-to-face trust, licensing, and real-world accountability remain comparatively resilient in the jobs...
- Referenced news from MIT Technology Review: Three things in AI to watch, according to a Nobel-winning economist https://www.technologyreview.com/2026/05/11/1137090/three-things-in-ai-to-watch-according-to-a-nobel-winning-economist/
- Referenced news from MIT Technology Review: Fostering breakthrough AI innovation through customer-back engineering https://www.technologyreview.com/2026/05/11/1136967/fostering-breakthrough-ai-innovation-through-customer-back-engineering/
- Referenced news from MIT Technology Review: Implementing advanced AI technologies in finance https://www.technologyreview.com/2026/05/11/1136786/implementing-advanced-ai-technologies-in-finance/
- Referenced news from Wired: I Work in Hollywood. Everyone Who Used to Make TV Is Now Secretly Training AI https://www.wired.com/story/i-work-in-hollywood-everyone-who-used-to-make-tv-now-training-ai/
- Referenced news from Wired: CUDA Proves Nvidia Is a Software Company https://www.wired.com/story/cuda-proves-nvidia-is-a-software-company/
- Referenced news from Wired: Hackable Robot Lawn Mower Unlocks a New Nightmare https://www.wired.com/story/security-news-this-week-hackable-robot-lawnmower-unlocks-a-new-nightmare/
- Referenced news from MIT Technology Review: Musk v. Altman week 2: OpenAI fires back, and Shivon Zilis reveals that Musk tried to poach Sam Altman https://www.technologyreview.com/2026/05/08/1137008/musk-v-altman-week-2-openai-fires-back-and-shivon-zilis-reveals-that-musk-tried-to-poach-sam-altman/
- Referenced news from Wired: There's a Long Shot Proposal to Protect California Workers From AI https://www.wired.com/story/tom-steyer-proposes-jobs-guarantee-to-protect-california-workers-from-ai/
- Referenced news from Wired: Nick Bostrom Has a Plan for Humanity’s ‘Big Retirement’ https://www.wired.com/story/nick-bostrom-has-a-plan-for-humanitys-big-retirement/
- Referenced news from Wired: The New Wild West of AI Kids’ Toys https://www.wired.com/story/the-new-wild-west-of-ai-kids-toys/
- Referenced news from Wired: Musk vs. Altman Evidence Shows What Microsoft Executives Thought of OpenAI https://www.wired.com/story/microsoft-executives-discuss-openai-sam-altman-2018/
- Referenced news from Wired: Trump Pivots on AI Regulation, Worker Ousted by DOGE Runs for Office, and Hantavirus Explained https://www.wired.com/story/uncanny-valley-podcast-trump-pivots-ai-regulation-worker-ousted-by-doge-runs-for-office-hantavirus-explained/
- Referenced news from Wired: How to Disable Google's Gemini in Chrome https://www.wired.com/story/you-can-disable-gemini-in-chrome-if-its-freaking-you-out/
- Referenced news from Wired: ChatGPT Has 'Goblin' Mania in the US. In China It Will 'Catch You Steadily' https://www.wired.com/story/chatgpt-chinese-catch-you-steadily-sycophancy/
- Referenced news from Wired: This Reggae Band Is in a Nightmare Battle Against AI Slop Remixes https://www.wired.com/story/this-reggae-band-is-in-a-nightmare-battle-against-ai-slop-remixes/
- Referenced news from Wired: Thousands of Vibe-Coded Apps Expose Corporate and Personal Data on the Open Web https://www.wired.com/story/thousands-of-vibe-coded-apps-expose-corporate-and-personal-data-on-the-open-web/
- Referenced news from Wired: Elon Musk’s Last-Ditch Effort to Control OpenAI: Recruit Sam Altman to Tesla https://www.wired.com/story/elon-musk-recruit-sam-altman-tesla-ai-lab-trial/
- Referenced news from Wired: Anthropic Gets in Bed With SpaceX as the AI Race Turns Weird https://www.wired.com/story/anthropic-spacex-compute-deal-colossus/
- Referenced news from Wired: Using AI for Just 10 Minutes Might Make You Lazy and Dumb, Study Shows https://www.wired.com/story/using-ai-negative-impact-thinking-problem-solving-study/
- Referenced news from Wired: I Am Begging AI Companies to Stop Naming Features After Human Processes https://www.wired.com/story/i-am-begging-ai-companies-to-stop-naming-features-after-human-processes/

### 2026-05-06

- URL: https://ai-job-risk.net/weekly-summaries/2026-05-06
- Summary: This week’s AI job risk update is mostly stable, with only small relative changes across occupations. The biggest signals came from stronger enterprise AI adoption, faster hardware demand, new model-debugging tools, and several security and autonomy stories. Apple’s comment that AI adoption is happening faster than expected, plus discussion of operationalizing AI for scale and sovereignty, supports slightly higher risk for digital knowledge-work roles tied to content, analysis, support, and software production. Goodfire’s new mechanistic interpretability tool also modestly strengthens the outlook for more controllable and deployable models in coding and information-heavy workflows. At the same time, cybersecurity news—including NSA testing Anthropic tools, OpenAI’s advanced security mode, and broader warnings about cyber-insecurity in the AI era—reinforces demand for human security oversight, tempering “jobs AI will replace” narratives for some defense-oriented roles. Waymo complaints and ongoing limits in robotics and real-world autonomy continue to support lower near-term AI replacement odds for many physical, safety-critical, and hands-on jobs often seen as more AI-proof jobs.
- Referenced news from MIT Technology Review: Week one of the Musk v. Altman trial: What it was like in the room https://www.technologyreview.com/2026/05/04/1136826/week-one-of-the-musk-v-altman-trial-what-it-was-like-in-the-room/
- Referenced news from Wired: Disneyland Now Uses Face Recognition on Visitors https://www.wired.com/story/security-news-this-week-disneyland-now-uses-face-recognition-on-visitors/
- Referenced news from MIT Technology Review: Musk v. Altman week 1: Elon Musk says he was duped, warns AI could kill us all, and admits that xAI distills OpenAI’s models https://www.technologyreview.com/2026/05/01/1136800/musk-v-altman-week-1-musk-says-he-was-duped-warns-ai-could-kill-us-all-and-admits-that-xai-distills-openais-models/
- Referenced news from Wired: A Dark-Money Campaign Is Paying Influencers to Frame Chinese AI as a Threat https://www.wired.com/story/super-pac-backed-by-openai-and-palantir-is-paying-tiktok-influencers-to-fear-monger-about-china/
- Referenced news from MIT Technology Review: Cyber-Insecurity in the AI Era https://www.technologyreview.com/2026/05/01/1136779/cyber-insecurity-in-the-ai-era/
- Referenced news from MIT Technology Review: Operationalizing AI for Scale and Sovereignty https://www.technologyreview.com/2026/05/01/1136772/operationalizing-ai-for-scale-and-sovereignty/
- Referenced news from MIT Technology Review: A new T-Mobile network for Christians aims to block porn and gender-related content https://www.technologyreview.com/2026/05/01/1136739/a-new-t-mobile-network-for-christians-aims-to-block-porn-and-gender-related-content/
- Referenced news from Wired: How Shivon Zilis Operated as Elon Musk’s OpenAI Insider https://www.wired.com/story/model-behavior-why-everything-in-musk-v-altman-leads-back-to-shivon-zelis/
- Referenced news from Wired: Good Luck Getting a Mac Mini for the Next ‘Several Months’ https://www.wired.com/story/apple-sold-out-mac-mini-openclaw/
- Referenced news from Wired: Musk v. Altman Kicks Off, DOJ Guts Voting Rights Unit, and Is the AI Job Apocalypse Overhyped? https://www.wired.com/story/uncanny-valley-podcast-musk-v-altman-doj-guts-voting-rights-unit-is-ai-job-apocalypse-overhyped/
- Referenced news from Wired: Elon Musk Seemingly Admits xAI Has Used OpenAI's Models to Train Its Own https://www.wired.com/story/elon-musk-distill-openai-models-partly-xai/
- Referenced news from Wired: OpenAI Rolls Out ‘Advanced’ Security Mode for At-Risk Accounts https://www.wired.com/story/openai-chatgpt-codex-advanced-account-security/
- Referenced news from MIT Technology Review: This startup’s new mechanistic interpretability tool lets you debug LLMs https://www.technologyreview.com/2026/04/30/1136721/this-startups-new-mechanistic-interpretability-tool-lets-you-debug-llms/
- Referenced news from Wired: These Men Allegedly Profit Off Teaching People How to Make AI Porn https://www.wired.com/story/ai-porn-lawsuit-arizona/
- Referenced news from Wired: Reid Hoffman Thinks Doctors Should Ask AI for a Second Opinion https://www.wired.com/story/reid-hoffman-ai-doctor-second-opinion-wired-health/
- Referenced news from Wired: How Elon Musk Squeezed OpenAI: They 'Are Gonna Want to Kill Me’ https://www.wired.com/story/model-behavior-elon-musk-cross-examined-sam-altman/
- Referenced news from Wired: Emergency First Responders Say Waymos Are Getting Worse https://www.wired.com/story/emergency-first-responders-say-waymos-are-getting-worse/
- Referenced news from Wired: Taylor Swift Wants to Trademark Her Likeness. These TikTok Deepfake Ads Show Why https://www.wired.com/story/taylor-swift-rihanna-tiktok-deepfake-ads/
- Referenced news from Wired: Sanctioned Chinese AI Firm SenseTime Releases Image Model Built for Speed https://www.wired.com/story/chinese-ai-giant-sensetime-is-running-its-new-model-on-chinese-chips/
- Referenced news from Wired: When Robots Have Their ChatGPT Moment, Remember These Pincers https://www.wired.com/story/when-robots-have-their-chatgpt-moment-remember-these-pincers/

### 2026-04-29

- URL: https://ai-job-risk.net/weekly-summaries/2026-04-29
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations. The clearest signal for jobs at risk from AI came from stronger general-purpose models and wider enterprise deployment: DeepSeek V4’s longer-context, open-source release and ongoing adoption of copilots and agents in finance, HR, supply chains, and customer operations raise pressure on text-heavy, repetitive knowledge work. That slightly increases risk for roles such as programmers, data analysts, copywriters, translators, and some support functions. At the same time, several stories point to limits on near-term replacement: enterprises still face major data-stack bottlenecks, and advice-heavy uses like financial guidance remain error-prone, which tempers risk for accountants, financial analysts, and other high-accountability professions. AI progress in science and drug discovery supports task augmentation more than full substitution for researchers and laboratory roles. Cyber misuse stories also increase demand for human oversight in cybersecurity. Overall, the latest news reinforces familiar patterns in jobs AI will replace first—routine digital work—while many hands-on, licensed,...
- Referenced news from MIT Technology Review: The missing step between hype and profit https://www.technologyreview.com/2026/04/27/1136456/the-missing-step-between-hype-and-profit/
- Referenced news from Wired: The Man Behind AlphaGo Thinks AI Is Taking the Wrong Path https://www.wired.com/story/david-silver-ai-ineffable-intelligence-reinforcement-learning/
- Referenced news from MIT Technology Review: Rebuilding the data stack for AI https://www.technologyreview.com/2026/04/27/1136322/rebuilding-the-data-stack-for-ai/
- Referenced news from Wired: Discord Sleuths Gained Unauthorized Access to Anthropic’s Mythos https://www.wired.com/story/security-news-this-week-discord-sleuths-gained-unauthorized-access-to-anthropics-mythos/
- Referenced news from Wired: Ace the Ping-Pong Robot Can Whup Your Ass https://www.wired.com/story/ace-the-robot-wants-to-become-the-world-table-tennis-champion/
- Referenced news from MIT Technology Review: Three reasons why DeepSeek’s new model V4 matters https://www.technologyreview.com/2026/04/24/1136422/why-deepseeks-v4-matters/
- Referenced news from Wired: AI-Designed Drugs by a DeepMind Spinoff Are Headed to Human Trials https://www.wired.com/story/wired-health-2026-how-ai-is-powering-drug-discovery-max-jaderberg/
- Referenced news from Wired: Apple's Next CEO Needs to Launch a Killer AI Product https://www.wired.com/story/apples-next-ceo-needs-to-launch-a-killer-ai-product/
- Referenced news from Wired: The Men Behind Your Favorite AI Gay Thirst Traps https://www.wired.com/story/gay-ai-instagram-influencers-red-carpet/
- Referenced news from Wired: 5 Reasons to Think Twice Before Using ChatGPT—or Any Chatbot—for Financial Advice https://www.wired.com/story/5-reasons-to-think-twice-before-using-chatgpt-for-financial-advice/
- Referenced news from Wired: Apple’s Next Chapter, SpaceX and Cursor Strike a Deal, and Palantir’s Controversial Manifesto https://www.wired.com/story/uncanny-valley-podcast-apple-next-chapter-spacex-cursor-deal-palantir-manifesto/
- Referenced news from Wired: At 'AI Coachella,' Stanford Students Line Up to Learn From Silicon Valley Royalty https://www.wired.com/story/stanford-cs-class-ai-coachella-ben-horowitz/
- Referenced news from Wired: Sam Altman’s Orb Company Promoted a Bruno Mars Partnership That Doesn't Exist https://www.wired.com/story/sam-altman-orb-company-bruno-mars-partnership-fake/
- Referenced news from Wired: 5 AI Models Tried to Scam Me. Some of Them Were Scary Good https://www.wired.com/story/ai-model-phishing-attack-cybersecurity/
- Referenced news from Wired: AI Tools Are Helping Mediocre North Korean Hackers Steal Millions https://www.wired.com/story/ai-tools-are-helping-mediocre-north-korean-hackers-steal-millions/
- Referenced news from MIT Technology Review: AI needs a strong data fabric to deliver business value https://www.technologyreview.com/2026/04/22/1135295/ai-needs-a-strong-data-fabric-to-deliver-business-value/
- Referenced news from Wired: Join Our Livestream: Musk v. Altman and the Future of OpenAI https://www.wired.com/story/livestream-musk-v-altman-trial/
- Referenced news from Wired: The Pope’s Warnings About AI Were AI-Generated, a Detection Tool Claims https://www.wired.com/story/pope-tweets-ai-generated-pangram-chrome-extension/
- Referenced news from MIT Technology Review: Resistance https://www.technologyreview.com/2026/04/21/1135665/resistance-ai-artificial-intelligence-backlash-protests/
- Referenced news from MIT Technology Review: Artificial scientists https://www.technologyreview.com/2026/04/21/1135663/artificial-scientists-ai-artificial-intelligence/

### 2026-04-22

- URL: https://ai-job-risk.net/weekly-summaries/2026-04-22
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations. The clearest signals in jobs at risk from AI came from enterprise deployment news: Chinese tech workers reportedly being asked to train AI doubles, Google expanding AI Mode in Chrome, and new funding and office expansion from the UK sovereign AI fund and Anthropic in London. Those developments reinforce near-term pressure on digital, repeatable knowledge work such as support, writing, software, and search-dependent marketing tasks. Hardware and robotics news, including Schematik’s design tooling and broader reporting on how robots learn, nudged a few engineering and drafting roles slightly higher, but physical field work remains far more AI-proof jobs territory. At the same time, public pushback from journalists against AI drafting and continued governance constraints in government and regulated environments kept some human-trust roles from rising further. Overall, the latest pattern in jobs AI will replace remains concentrated in clerical, content, support, and standardized analytical work rather than hands-on, high-liability, or relationship-intensive occupations.
- Referenced news from Wired: Prego Has a Dinner-Conversation-Recording Device, Capisce? https://www.wired.com/story/prego-has-a-dinner-recording-device-capiche/
- Referenced news from Wired: Tech CEOs Think AI Will Let Them Be Everywhere at Once https://www.wired.com/story/tech-ceos-using-ai-to-be-everywhere-at-once/
- Referenced news from MIT Technology Review: Chinese tech workers are starting to train their AI doubles–and pushing back https://www.technologyreview.com/2026/04/20/1136149/chinese-tech-workers-ai-colleagues/
- Referenced news from Wired: It Takes 2 Minutes to Hack the EU’s New Age-Verification App https://www.wired.com/story/security-news-this-week-it-takes-2-minutes-to-hack-the-eus-new-age-verification-app/
- Referenced news from Wired: Schematik Is ‘Cursor for Hardware.’ Anthropic Wants In https://www.wired.com/story/schematik-is-cursor-for-hardware-anthropic-wants-in-on-it/
- Referenced news from Wired: OpenAI Executive Kevin Weil Is Leaving the Company https://www.wired.com/story/openai-executive-kevin-weil-is-leaving-the-company/
- Referenced news from Wired: Gazing Into Sam Altman’s Orb Now Proves You’re Human on Tinder https://www.wired.com/story/gazing-into-sam-altmans-orb-now-proves-youre-human-on-tinder/
- Referenced news from Wired: AI Drafting My Stories? Over My Dead Body https://www.wired.com/story/backchannel-the-problem-with-letting-ai-do-the-writing/
- Referenced news from MIT Technology Review: How robots learn: A brief, contemporary history https://www.technologyreview.com/2026/04/17/1135416/how-robots-learn-brief-contemporary-history/
- Referenced news from Wired: The Battle for OpenAI’s Soul https://www.wired.com/story/musk-v-altman-trial-openai-xai/
- Referenced news from Wired: The UK Launches Its $675 Million Sovereign AI Fund https://www.wired.com/story/the-uk-launches-its-dollar675-million-sovereign-ai-fund/
- Referenced news from Wired: Google's AI Mode Update Tries to Kill Tab Hopping in Chrome https://www.wired.com/story/google-ai-mode-update-tries-to-kill-tab-hopping-in-chrome/
- Referenced news from Wired: Anthropic Plots Major London Expansion https://www.wired.com/story/anthropic-plots-major-london-expansion/
- Referenced news from MIT Technology Review: Treating enterprise AI as an operating layer https://www.technologyreview.com/2026/04/16/1135554/treating-enterprise-ai-as-an-operating-layer/
- Referenced news from MIT Technology Review: Making AI operational in constrained public sector environments https://www.technologyreview.com/2026/04/16/1135216/making-ai-operational-in-constrained-public-sector-environments/
- Referenced news from MIT Technology Review: Why having “humans in the loop” in an AI war is an illusion https://www.technologyreview.com/2026/04/16/1136029/humans-in-the-loop-ai-war-illusion/
- Referenced news from Wired: This Beanie Is Designed to Read Your Thoughts https://www.wired.com/story/this-beanie-is-designed-to-read-your-thoughts/
- Referenced news from Wired: AI Could Democratize One of Tech's Most Valuable Resources https://www.wired.com/story/ai-could-democratize-one-of-techs-most-valuable-resources/
- Referenced news from Wired: Allbirds Is Pivoting to AI Compute. Sure, Why Not https://www.wired.com/story/allbirds-is-pivoting-to-ai-compute-sure-why-not/
- Referenced news from Wired: AI Slop Is Making the Internet Fake-Happy https://www.wired.com/story/ai-slop-is-changing-the-internet-just-not-how-you-might-think/

### 2026-04-15

- URL: https://ai-job-risk.net/weekly-summaries/2026-04-15
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations. The biggest signals came from continued progress in AI agents, including Microsoft’s new OpenClaw-like work on task-completing agents and strong enterprise momentum around Claude and AI-powered developer tools. That slightly raises exposure for jobs built around repeatable digital workflows, software support, and structured knowledge work. At the same time, cybersecurity and governance-heavy roles look a bit more resilient because Anthropic’s Mythos-related security concerns, facial-recognition backlash, and new lawsuits around harmful AI use all reinforce the need for human oversight, compliance, and incident response. Early humanoid robot commercialization from Unitree is notable, but it still looks too immature to materially change near-term replacement odds for most physical jobs. Overall, the latest pattern in jobs AI will replace and jobs at risk from AI still favors higher pressure on clerical, support, and routine content tasks, while many AI-proof jobs remain those requiring trust, dexterity, regulation, or high-stakes human judgment.
- Referenced news from Wired: You Can Soon Buy a $4,370 Humanoid Robot on AliExpress https://www.wired.com/story/unitree-r1-humanoid-robot-for-sale-on-aliexpress/
- Referenced news from TechCrunch: Microsoft is working on yet another OpenClaw-like agent https://techcrunch.com/2026/04/13/microsoft-is-working-on-yet-another-openclaw-like-agent/
- Referenced news from TechCrunch: Stanford report highlights growing disconnect between AI insiders and everyone else https://techcrunch.com/2026/04/13/stanford-report-highlights-growing-disconnect-between-ai-insiders-and-everyone-else/
- Referenced news from Wired: Meta Is Warned That Facial Recognition Glasses Will Arm Sexual Predators https://www.wired.com/story/meta-ray-ban-oakley-smart-glasses-no-face-recognition-civil-society/
- Referenced news from MIT Technology Review: Why opinion on AI is so divided https://www.technologyreview.com/2026/04/13/1135720/why-opinion-on-ai-is-so-divided/
- Referenced news from TechCrunch: Vercel CEO Guillermo Rauch signals IPO readiness as AI agents fuel revenue surge https://techcrunch.com/2026/04/13/vercel-ceo-guillermo-rauch-signals-ipo-readiness-as-ai-agents-fuel-revenue-surge/
- Referenced news from MIT Technology Review: Want to understand the current state of AI? Check out these charts. https://www.technologyreview.com/2026/04/13/1135675/want-to-understand-the-current-state-of-ai-check-out-these-charts/
- Referenced news from Wired: The Internet's Most Powerful Archiving Tool Is in Peril https://www.wired.com/story/the-internets-most-powerful-archiving-tool-is-in-mortal-peril/
- Referenced news from Wired: AI Agents Are Coming for Your Dating Life https://www.wired.com/story/ai-agents-are-coming-for-your-dating-life-next/
- Referenced news from TechCrunch: The largest orbital compute cluster is open for business https://techcrunch.com/2026/04/13/the-largest-orbital-compute-cluster-is-open-for-business/
- Referenced news from TechCrunch: Trump officials may be encouraging banks to test Anthropic’s Mythos model https://techcrunch.com/2026/04/12/trump-officials-may-be-encouraging-banks-to-test-anthropics-mythos-model/
- Referenced news from TechCrunch: Apple reportedly testing four designs for upcoming smart glasses https://techcrunch.com/2026/04/12/apple-reportedly-testing-four-designs-for-upcoming-smart-glasses/
- Referenced news from TechCrunch: From LLMs to hallucinations, here’s a simple guide to common AI terms https://techcrunch.com/2026/04/12/artificial-intelligence-definition-glossary-hallucinations-guide-to-common-ai-terms/
- Referenced news from TechCrunch: At the HumanX conference, everyone was talking about Claude https://techcrunch.com/2026/04/12/at-the-humanx-conference-everyone-was-talking-about-claude/
- Referenced news from TechCrunch: Sam Altman responds to ‘incendiary’ New Yorker article after attack on his home https://techcrunch.com/2026/04/11/sam-altman-responds-to-incendiary-new-yorker-article-after-attack-on-his-home/
- Referenced news from Wired: Your Push Notifications Aren’t Safe From the FBI https://www.wired.com/story/security-news-this-week-your-push-notifications-arent-safe-from-the-fbi/
- Referenced news from Wired: How the Internet Broke Everyone’s Bullshit Detectors https://www.wired.com/story/how-the-internet-broke-everyones-bullshit-detectors/
- Referenced news from TechCrunch: Anthropic temporarily banned OpenClaw’s creator from accessing Claude https://techcrunch.com/2026/04/10/anthropic-temporarily-banned-openclaws-creator-from-accessing-claude/
- Referenced news from Wired: Anthropic’s Mythos Will Force a Cybersecurity Reckoning—Just Not the One You Think https://www.wired.com/story/anthropics-mythos-will-force-a-cybersecurity-reckoning-just-not-the-one-you-think/
- Referenced news from TechCrunch: Stalking victim sues OpenAI, claims ChatGPT fueled her abuser’s delusions and ignored her warnings https://techcrunch.com/2026/04/10/stalking-victim-sues-openai-claims-chatgpt-fueled-her-abusers-delusions-and-ignored-her-warnings/

### 2026-04-08

- URL: https://ai-job-risk.net/weekly-summaries/2026-04-08
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations. The clearest signal is broader deployment of AI assistants through new ChatGPT app integrations with tools like Canva, Figma, Spotify, Expedia, Uber, and DoorDash, which strengthens AI’s practical reach into digital workflows rather than proving immediate full job replacement. That raises exposure slightly for some jobs at risk from AI in content, design, support, scheduling, and software tasks. At the same time, Microsoft’s warning that Copilot is “for entertainment purposes only,” plus ongoing security incidents around AI tooling and code ecosystems, reinforces limits on high-trust autonomous use, slightly reducing near-term AI job risk for some analytical, legal, and engineering roles. Physical-world jobs remain comparatively more AI-proof jobs this week, though Japan’s push to deploy robots where labor is scarce nudges a few operational roles higher at the margin. Overall, the latest news suggests gradual workflow substitution, not a sudden jump in which jobs AI will replace.
- Referenced news from MIT Technology Review: The one piece of data that could actually shed light on your job and AI https://www.technologyreview.com/2026/04/06/1135187/the-one-piece-of-data-that-could-actually-shed-light-on-your-job-and-ai/
- Referenced news from TechCrunch: OpenAI’s vision for the AI economy: public wealth funds, robot taxes, and a four-day work week https://techcrunch.com/2026/04/06/openais-vision-for-the-ai-economy-public-wealth-funds-robot-taxes-and-a-four-day-work-week/
- Referenced news from TechCrunch: Startup Battlefield 200 applications open: A chance for VC access, TechCrunch coverage, and $100K https://techcrunch.com/2026/04/06/startup-battlefield-200-applications-open-get-vc-access-techcrunch-coverage-and-100k/
- Referenced news from TechCrunch: How to use the new ChatGPT app integrations, including DoorDash, Spotify, Uber, and others https://techcrunch.com/2026/04/06/how-to-use-chatgpt-apps-doordash-spotify-uber/
- Referenced news from TechCrunch: Ticket savings of up to $500 this week for TechCrunch Disrupt 2026 https://techcrunch.com/2026/04/06/massive-ticket-savings-of-up-to-500-this-week-for-techcrunch-disrupt-2026/
- Referenced news from TechCrunch: Spain’s Xoople raises $130 million Series B to map the Earth for AI https://techcrunch.com/2026/04/06/spains-xoople-raises-130-million-series-b-to-map-the-earth-for-ai/
- Referenced news from MIT Technology Review: AI is changing how small online sellers decide what to make https://www.technologyreview.com/2026/04/06/1135118/ai-online-seller-alibaba-accio/
- Referenced news from Wired: The Ridiculously Nerdy Intel Bet That Could Rake in Billions https://www.wired.com/story/why-chip-packaging-could-decide-the-next-phase-of-the-ai-boom/
- Referenced news from TechCrunch: Copilot is ‘for entertainment purposes only,’ according to Microsoft’s terms of service https://techcrunch.com/2026/04/05/copilot-is-for-entertainment-purposes-only-according-to-microsofts-terms-of-service/
- Referenced news from TechCrunch: Can orbital data centers help justify a massive valuation for SpaceX? https://techcrunch.com/2026/04/05/can-orbital-data-centers-help-justify-a-massive-valuation-for-spacex/
- Referenced news from TechCrunch: In Japan, the robot isn’t coming for your job; it’s filling the one nobody wants https://techcrunch.com/2026/04/05/japan-is-proving-experimental-physical-ai-is-ready-for-the-real-world/
- Referenced news from TechCrunch: Anthropic says Claude Code subscribers will need to pay extra for OpenClaw support https://techcrunch.com/2026/04/04/anthropic-says-claude-code-subscribers-will-need-to-pay-extra-for-openclaw-support/
- Referenced news from Wired: Hackers Are Posting the Claude Code Leak With Bonus Malware https://www.wired.com/story/security-news-this-week-hackers-are-posting-the-claude-code-leak-with-bonus-malware/
- Referenced news from TechCrunch: Anthropic is having a moment in the private markets; SpaceX could spoil the party https://techcrunch.com/2026/04/03/anthropic-is-having-a-moment-in-the-private-markets-spacex-could-spoil-the-party/
- Referenced news from Wired: Meta Pauses Work With Mercor After Data Breach Puts AI Industry Secrets at Risk https://www.wired.com/story/meta-pauses-work-with-mercor-after-data-breach-puts-ai-industry-secrets-at-risk/
- Referenced news from TechCrunch: OpenAI executive shuffle includes new role for COO Brad Lightcap to lead ‘special projects’ https://techcrunch.com/2026/04/03/openai-executive-shuffle-new-roles-coo-brad-lightcap-fidji-simo-kate-rouch/
- Referenced news from TechCrunch: Anthropic buys biotech startup Coefficient Bio in $400M deal: reports https://techcrunch.com/2026/04/03/anthropic-buys-biotech-startup-coefficient-bio-in-400m-deal-reports/
- Referenced news from TechCrunch: Anthropic ramps up its political activities with a new PAC https://techcrunch.com/2026/04/03/anthropic-ramps-up-its-political-activities-with-a-new-pac/
- Referenced news from TechCrunch: AI companies are building huge natural gas plants to power data centers. What could go wrong? https://techcrunch.com/2026/04/03/ai-companies-are-building-huge-natural-gas-plants-to-power-data-centers-what-could-go-wrong/
- Referenced news from TechCrunch: AI companies are building huge natural gas plants to power data centers. What could go wrong? https://techcrunch.com/2026/04/03/ai-energy-microsoft-meta-google-natural-gas-mining-fomo/

### 2026-04-01

- URL: https://ai-job-risk.net/weekly-summaries/2026-04-01
- Summary: This week’s AI job risk update is mostly stable, with only small relative adjustments as the latest AI news was mixed. The biggest signal raising exposure for some knowledge-work roles is stronger mainstream adoption of general-purpose assistants: Anthropic said Claude paid subscriptions have more than doubled this year, and Google launched easier switching tools into Gemini. That supports slightly higher near-term risk for jobs at risk from AI where core tasks are text-heavy, research-heavy, or chatbot-compatible, such as translation, customer support, and administrative work. At the same time, OpenAI’s shutdown of Sora looks like a reality check for AI video and creative automation, limiting the case for sharp increases in video-first and some visual content roles. Stanford’s study on the dangers of chatbots giving personal advice also reinforces limits for counseling and therapy-adjacent work, supporting the idea that many interpersonal and regulated occupations remain relatively AI-proof jobs for now. Overall, the ranking of jobs AI will replace changes only modestly this week.
- Referenced news from TechCrunch: Why OpenAI really shut down Sora https://techcrunch.com/2026/03/29/why-openai-really-shut-down-sora/
- Referenced news from TechCrunch: Sora’s shutdown could be a reality check moment for AI video https://techcrunch.com/2026/03/29/soras-shutdown-could-be-a-reality-check-moment-for-ai-video/
- Referenced news from TechCrunch: Bluesky leans into AI with Attie, an app for building custom feeds https://techcrunch.com/2026/03/28/bluesky-leans-into-ai-with-attie-an-app-for-building-custom-feeds/
- Referenced news from TechCrunch: Stanford study outlines dangers of asking AI chatbots for personal advice https://techcrunch.com/2026/03/28/stanford-study-outlines-dangers-of-asking-ai-chatbots-for-personal-advice/
- Referenced news from TechCrunch: Elon Musk’s last co-founder reportedly leaves xAI https://techcrunch.com/2026/03/28/elon-musks-last-co-founder-reportedly-leaves-xai/
- Referenced news from TechCrunch: Anthropic’s Claude popularity with paying consumers is skyrocketing https://techcrunch.com/2026/03/28/anthropics-claude-popularity-with-paying-consumers-is-skyrocketing/
- Referenced news from Wired: AI Research Is Getting Harder to Separate From Geopolitics https://www.wired.com/story/made-in-china-ai-research-is-starting-to-split-along-geopolitical-lines/
- Referenced news from TechCrunch: Why SoftBank’s new $40B loan points to a 2026 OpenAI IPO https://techcrunch.com/2026/03/27/why-softbanks-new-40b-loan-points-to-a-2026-openai-ipo/
- Referenced news from TechCrunch: Memory chip giant SK hynix could help end ‘RAMmageddon’ with blockbuster US IPO https://techcrunch.com/2026/03/27/memory-chip-giant-sk-hynix-could-help-end-rammageddon-with-blockbuster-us-ipo/
- Referenced news from TechCrunch: VCs are betting billions on AI’s next wave, so why is OpenAI killing Sora? https://techcrunch.com/podcast/vcs-are-betting-billions-on-ais-next-wave-so-why-is-openai-killing-sora/
- Referenced news from Wired: Apple Still Plans to Sell iPhones When It Turns 100 https://www.wired.com/story/apple-50-year-anniversary-artificial-intelligence-iphone/
- Referenced news from TechCrunch: OpenAI shuts down Sora while Meta gets shut out in court https://techcrunch.com/video/openai-shuts-down-sora-while-meta-gets-shut-out-in-court/
- Referenced news from Wired: A New AI Documentary Puts CEOs in the Hot Seat—but Goes Too Easy on Them https://www.wired.com/story/a-new-ai-documentary-puts-ceos-in-the-hot-seat-but-goes-too-easy-on-them/
- Referenced news from Wired: I Asked ChatGPT 500 Questions. Here Are the Ads I Saw Most Often https://www.wired.com/story/i-asked-chatgpt-500-questions-here-are-the-ads-i-saw-most-often/
- Referenced news from TechCrunch: David Sacks is done as AI czar — here’s what he’s doing instead https://techcrunch.com/2026/03/26/david-sacks-is-done-as-ai-czar-heres-what-hes-doing-instead/
- Referenced news from TechCrunch: Anthropic wins injunction against Trump administration over Defense Department saga https://techcrunch.com/2026/03/26/anthropic-wins-injunction-against-trump-administration-over-defense-department-saga/
- Referenced news from TechCrunch: You can now transfer your chats and personal information from other chatbots directly into Gemini https://techcrunch.com/2026/03/26/you-can-now-transfer-your-chats-and-personal-information-from-other-chatbots-directly-into-gemini/
- Referenced news from Wired: Anthropic Supply-Chain Risk Designation Halted By Judge https://www.wired.com/story/anthropic-supply-chain-risk-designation-injunction/
- Referenced news from TechCrunch: Wikipedia cracks down on the use of AI in article writing https://techcrunch.com/2026/03/26/wikipedia-cracks-down-on-the-use-of-ai-in-article-writing/
- Referenced news from TechCrunch: OpenAI abandons yet another side quest: ChatGPT’s erotic mode https://techcrunch.com/2026/03/26/openai-abandons-yet-another-side-quest-chatgpts-erotic-mode/

### 2026-03-25

- URL: https://ai-job-risk.net/weekly-summaries/2026-03-25
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across professions. The strongest signals came from better AI infrastructure and workflow automation: Gimlet Labs’ multi-chip inference push, Amazon’s Trainium momentum, and Littlebird’s real-time screen-reading assistant all improve the practicality of AI for office software, admin support, data handling, coding, and digital content tasks. That slightly raises exposure for some of the jobs at risk from AI in clerical, support, and software-adjacent work. At the same time, several headlines reinforced limits on full replacement: AI delusion concerns, compliance controversy, and defense-sector governance disputes highlight reliability, oversight, and accountability gaps. Those constraints slightly lower near-term replacement risk for trust-heavy roles like auditors, cybersecurity analysts, and lawyers. Overall, the ranking of jobs AI will replace versus more AI-proof jobs changed only modestly this week: routine screen-based work edged up, while regulated, judgment-intensive, and high-accountability roles held firmer.
- Referenced news from MIT Technology Review: The hardest question to answer about AI-fueled delusions https://www.technologyreview.com/2026/03/23/1134527/the-hardest-question-to-answer-about-ai-fueled-delusions/
- Referenced news from TechCrunch: Startup Gimlet Labs is solving the AI inference bottleneck in a surprisingly elegant way https://techcrunch.com/2026/03/23/startup-gimlet-labs-is-solving-the-ai-inference-bottleneck-in-a-surprisingly-elegant-way/
- Referenced news from TechCrunch: Littlebird raises $11M for its AI-assisted ‘recall’ tool that reads your computer screen https://techcrunch.com/2026/03/23/littlebird-raises-11m-to-capture-context-from-your-computer-so-you-can-query-your-data/
- Referenced news from TechCrunch: Elizabeth Warren calls Pentagon’s decision to bar Anthropic ‘retaliation’ https://techcrunch.com/2026/03/23/elizabeth-warren-anthropic-pentagon-defense-supply-chain-risk-retaliation/
- Referenced news from TechCrunch: Sam Altman-backed fusion startup Helion in talks to sell power to OpenAI https://techcrunch.com/2026/03/23/sam-altman-openai-fusion-energy-board-helion/
- Referenced news from TechCrunch: Sam Altman-backed fusion startup Helion in talks with OpenAI https://techcrunch.com/2026/03/23/sam-altman-backed-fusion-startup-helion-in-talks-with-openai/
- Referenced news from Wired: Meet the Gods of AI Warfare https://www.wired.com/story/project-maven-katrina-manson-book-excerpt/
- Referenced news from MIT Technology Review: The Bay Area’s animal welfare movement wants to recruit AI https://www.technologyreview.com/2026/03/23/1134491/the-bay-areas-animal-welfare-movement-wants-to-recruit-ai/
- Referenced news from Wired: The AI Race Is Pressuring Utilities to Squeeze More From Europe’s Power Grids https://www.wired.com/story/europe-squeeze-power-energy-grid-ai-data-center/
- Referenced news from TechCrunch: Do you want to build a robot snowman? https://techcrunch.com/2026/03/22/do-you-want-to-build-a-robot-snowman/
- Referenced news from TechCrunch: Cursor admits its new coding model was built on top of Moonshot AI’s Kimi https://techcrunch.com/2026/03/22/cursor-admits-its-new-coding-model-was-built-on-top-of-moonshot-ais-kimi/
- Referenced news from TechCrunch: Elon Musk unveils chip manufacturing plans for SpaceX and Tesla https://techcrunch.com/2026/03/22/elon-musk-unveils-chip-manufacturing-plans-for-spacex-and-tesla/
- Referenced news from TechCrunch: Delve accused of misleading customers with ‘fake compliance’ https://techcrunch.com/2026/03/22/delve-accused-of-misleading-customers-with-fake-compliance/
- Referenced news from TechCrunch: An exclusive tour of Amazon’s Trainium lab, the chip that’s won over Anthropic, OpenAI, even Apple https://techcrunch.com/2026/03/22/an-exclusive-tour-of-amazons-trainium-lab-the-chip-thats-won-over-anthropic-openai-even-apple/
- Referenced news from TechCrunch: Are AI tokens the new signing bonus or just a cost of doing business? https://techcrunch.com/2026/03/21/are-ai-tokens-the-new-signing-bonus-or-just-a-cost-of-doing-business/
- Referenced news from TechCrunch: Publisher pulls horror novel ‘Shy Girl’ over AI concerns https://techcrunch.com/2026/03/21/publisher-pulls-horror-novel-shy-girl-over-ai-concerns/
- Referenced news from TechCrunch: Delve accused of misleading customers with ‘fake compliance’ https://techcrunch.com/2026/03/21/delve-accused-of-misleading-customers-with-fake-compliance/
- Referenced news from TechCrunch: Why Wall Street wasn’t won over by Nvidia’s big conference https://techcrunch.com/2026/03/21/why-wall-street-wasnt-won-over-by-nvidias-big-conference/
- Referenced news from Wired: I Tried DoorDash’s Tasks App and Saw the Bleak Future of AI Gig Work https://www.wired.com/story/i-tried-doordashs-tasks-app-and-saw-the-bleak-future-of-ai-gig-work/
- Referenced news from Wired: Cyberattack on a Car Breathalyzer Firm Leaves Drivers Stuck https://www.wired.com/story/security-news-this-week-cyberattack-on-a-car-breathalyzer-firm-leaves-drivers-stuck/

### 2026-03-18

- URL: https://ai-job-risk.net/weekly-summaries/2026-03-18
- Summary: This week’s AI job risk update is mostly stable, with only small relative moves across occupations as recent developments point more to deeper enterprise adoption than to sudden labor replacement. The clearest signal for jobs at risk from AI came from AI-native workflow software in lending and support functions: Fuse’s funding to modernize credit-union loan origination supports slightly higher risk for loan officers, underwriters, and adjacent clerical finance roles. ChatGPT app integrations with tools like Canva, Figma, Spotify, Uber, and Expedia also reinforce gradual automation pressure on digital coordination, content, and support tasks. At the same time, legal friction around AI training data, reported delays for ByteDance’s video generator, and safety concerns tied to chatbot harms modestly constrain the pace of replacement in some creative and advisory roles. Nvidia ecosystem momentum and agentic AI infrastructure continue to strengthen technical automation capabilities, but this week’s news does not justify broad score swings. Overall, the picture for jobs AI will replace versus AI-proof jobs remains one of incremental change, not sudden disruption.
- Referenced news from Wired: WIRED Article Production automation page/Only for QA/Do not click/Do not publish https://www.wired.com/story/wired-article-production-automation-page-only-for-qa-do-not-click-do-not-publish/
- Referenced news from TechCrunch: Another deep tech chip startup becomes a unicorn: Frore hits $1.64B https://techcrunch.com/2026/03/16/another-deep-tech-chip-startup-becomes-a-unicorn-frore-hits-1-64b/
- Referenced news from TechCrunch: Fuse raises $25M to disrupt aging loan origination systems used by U.S. credit unions https://techcrunch.com/2026/03/16/fuse-raises-25m-to-disrupt-aging-loan-origination-systems-used-by-u-s-credit-unions/
- Referenced news from TechCrunch: The dictionary sues OpenAI https://techcrunch.com/2026/03/16/merriam-webster-openai-encyclopedia-brittanica-lawsuit/
- Referenced news from TechCrunch: The dictionary sues OpenAI https://techcrunch.com/2026/03/16/the-dictionary-sues-openai/
- Referenced news from MIT Technology Review: Where OpenAI’s technology could show up in Iran https://www.technologyreview.com/2026/03/16/1134315/where-openais-technology-could-show-up-in-iran/
- Referenced news from TechCrunch: How to watch Jensen Huang’s Nvidia GTC 2026 keynote — and what to expect https://techcrunch.com/2026/03/16/nvidia-gtc-how-to-watch-jensen-huang-2026-keynote/
- Referenced news from MIT Technology Review: Nurturing agentic AI beyond the toddler stage https://www.technologyreview.com/2026/03/16/1133979/nurturing-agentic-ai-beyond-the-toddler-stage/
- Referenced news from MIT Technology Review: Securing digital assets against future threats https://www.technologyreview.com/2026/03/16/1134287/securing-digital-assets-against-future-threats/
- Referenced news from Wired: ‘100 Video Calls Per Day’: Models Are Applying to Be the Face of AI Scams https://www.wired.com/story/models-are-applying-to-be-the-face-of-ai-scams/
- Referenced news from TechCrunch: Google, Accel India accelerator choses 5 startups and none are ‘AI wrappers’ https://techcrunch.com/2026/03/15/google-and-accel-cut-through-wrappers-in-4000-ai-startup-pitches-to-pick-five-tied-to-india/
- Referenced news from TechCrunch: ByteDance reportedly pauses global launch of its Seedance 2.0 video generator https://techcrunch.com/2026/03/15/bytedance-reportedly-pauses-global-launch-of-its-seedance-2-0-video-generator/
- Referenced news from TechCrunch: Lawyer behind AI psychosis cases warns of mass casualty risks https://techcrunch.com/2026/03/15/lawyer-behind-ai-psychosis-cases-warns-of-mass-casualty-risks/
- Referenced news from TechCrunch: Wiz investor unpacks Google’s $32B acquisition https://techcrunch.com/2026/03/15/wiz-investor-unpacks-googles-32b-acquisition/
- Referenced news from TechCrunch: US Army announces contract with Anduril worth up to $20B https://techcrunch.com/2026/03/14/us-army-announces-contract-with-anduril-worth-up-to-20b/
- Referenced news from TechCrunch: Meta reportedly considering layoffs that could affect 20% of the company https://techcrunch.com/2026/03/14/meta-reportedly-considering-layoffs-that-could-affect-20-of-the-company/
- Referenced news from TechCrunch: How to use the new ChatGPT app integrations, including DoorDash, Spotify, Uber, and others https://techcrunch.com/2026/03/14/how-to-use-the-new-chatgpt-app-integrations-including-doordash-spotify-uber-and-others/
- Referenced news from TechCrunch: ‘Not built right the first time’ — Musk’s xAI is starting over again, again https://techcrunch.com/2026/03/13/not-built-right-the-first-time-musks-xai-is-starting-over-again-again/
- Referenced news from TechCrunch: Lawyer behind AI psychosis cases warns of mass casualty risks https://techcrunch.com/2026/03/13/lawyer-behind-ai-psychosis-cases-warns-of-mass-casualty-risks/
- Referenced news from TechCrunch: Nyne, founded by a father-son duo, gives AI agents the human context they’re missing https://techcrunch.com/2026/03/13/nyne-founded-by-a-father-son-duo-gives-ai-agents-the-human-context-theyre-missing/

### 2026-03-14

- URL: https://ai-job-risk.net/weekly-summaries/2026-03-14
- Summary: This week’s AI job risk update reflects faster rollout of AI agents in sales, customer messaging, and enterprise workflows—areas often cited in “jobs AI will replace” and “jobs at risk from AI” discussions. Rox AI’s $1.2B valuation signals strong demand for AI-native CRM and automated outreach, raising pressure on routine sales and coordination tasks. Meta AI replying to Facebook Marketplace buyers further normalizes automated customer communication, reinforcing displacement risk for front-line support, reception, and scheduling-heavy roles. In adjacent white-collar work, funding for Gumloop (AI agent builders) and Atlassian’s AI-driven staffing cuts highlight acceleration of internal automation across operations, IT, and documentation. Counterbalancing this, legal and policy scrutiny (e.g., the Grammarly lawsuit) suggests growing friction around deploying AI into editorial workflows, modestly supporting some human-heavy writing and editing roles. Overall AI job risk shifts remain small week-to-week to preserve relative ranking and identify more AI-proof jobs among hands-on, safety-critical roles.
- Referenced news from TechCrunch: Sales automation startup Rox AI hits $1.2B valuation, sources say https://techcrunch.com/2026/03/12/sales-automation-startup-rox-ai-hits-1-2b-valuation-sources-say/
- Referenced news from MIT Technology Review: Defense official reveals how AI chatbots could be used for targeting decisions https://www.technologyreview.com/2026/03/12/1134243/defense-official-military-use-ai-chatbots-targeting-decisions/
- Referenced news from Wired: ‘Uncanny Valley’: Anthropic’s DOD Lawsuit, War Memes, and AI Coming for VC Jobs https://www.wired.com/story/uncanny-valley-podcast-anthropic-department-defense-lawsuit-iran-war-memes-artificial-intelligence-venture-capital/
- Referenced news from TechCrunch: Facebook Marketplace now lets Meta AI respond to buyers’ messages https://techcrunch.com/2026/03/12/facebook-marketplace-now-lets-meta-ai-respond-to-buyers-messages/
- Referenced news from Wired: Google Is Not Ruling Out Ads in Gemini https://www.wired.com/story/google-nick-fox-advertising-search-ai-gemini/
- Referenced news from TechCrunch: Tinder tries to lure people back to online dating with IRL events, virtual speed dating https://techcrunch.com/2026/03/12/tinder-tries-to-lure-people-back-to-online-dating-with-irl-events-virtual-speed-dating/
- Referenced news from TechCrunch: Atlassian follows Block’s footsteps and cuts staff in the name of AI https://techcrunch.com/2026/03/12/atlassian-follows-blocks-footsteps-and-cuts-staff-in-the-name-of-ai/
- Referenced news from TechCrunch: Bumble introduces an AI dating assistant, ‘Bee’ https://techcrunch.com/2026/03/12/bumble-introduces-an-ai-dating-assistant-bee/
- Referenced news from TechCrunch: Bumble to launch an AI dating assistant, ‘Bee’ https://techcrunch.com/2026/03/12/bumble-to-launch-an-ai-dating-assistant-bee/
- Referenced news from TechCrunch: A writer is suing Grammarly for turning her and other authors into ‘AI editors’ without consent https://techcrunch.com/2026/03/12/a-writer-is-suing-grammarly-for-turning-her-and-other-authors-into-ai-editors-without-consent/
- Referenced news from TechCrunch: Gumloop lands $50M from Benchmark to turn every employee into an AI agent builder https://techcrunch.com/2026/03/12/gumloop-lands-50m-from-benchmark-to-turn-every-employee-into-an-ai-agent-builder/
- Referenced news from TechCrunch: Alexa+ gets a new ‘adults only’ personality option that curses but won’t do NSFW content https://techcrunch.com/2026/03/12/alexa-gets-a-new-adults-only-personality-option-that-curses-but-wont-do-nsfw-content/
- Referenced news from TechCrunch: Alexa+ gets a new ‘adults only’ personality option that curses but won’t get into NSFW content https://techcrunch.com/2026/03/12/alexa-gets-a-new-adults-only-personality-option-that-curses-but-wont-get-into-nsfw-content/
- Referenced news from TechCrunch: Wonderful raises $150M Series B at $2B valuation https://techcrunch.com/2026/03/12/wonderful-raises-150m-series-b-at-2b-valuation/
- Referenced news from MIT Technology Review: Pragmatic by design: Engineering AI for the real world https://www.technologyreview.com/2026/03/12/1133675/pragmatic-by-design-engineering-ai-for-the-real-world/
- Referenced news from Wired: Google Maps Gets Chatty With a New Gemini-Powered Interface https://www.wired.com/story/google-maps-ask-maps-gemini-powered-tool/
- Referenced news from TechCrunch: Google Maps is getting an AI ‘Ask Maps’ feature and upgraded ‘immersive’ navigation https://techcrunch.com/2026/03/12/google-maps-is-getting-an-ai-ask-maps-feature-and-upgraded-immersive-navigation/
- Referenced news from TechCrunch: Google is using old news reports and AI to predict flash floods https://techcrunch.com/2026/03/12/google-is-using-old-news-reports-and-ai-to-predict-flash-floods/
- Referenced news from TechCrunch: AI ‘actor’ Tilly Norwood put out the worst song I’ve ever heard https://techcrunch.com/2026/03/11/ai-actor-tilly-norwood-put-out-the-worst-song-ive-ever-heard/
- Referenced news from TechCrunch: Ford’s new AI assistant will help fleet owners know if seatbelts are being used https://techcrunch.com/2026/03/11/fords-new-ai-assistant-will-help-fleet-owners-know-if-seatbelts-are-being-used/

### 2026-03-05

- URL: https://ai-job-risk.net/weekly-summaries/2026-03-05
- Summary: This week’s AI job risk signals strengthened for several customer-facing and routine digital roles—often cited in “jobs AI will replace” and “jobs at risk from AI” lists—while some hands-on and regulated roles look slightly more AI‑proof in relative terms. The clearest labor-market development is Deutsche Telekom’s plan with ElevenLabs to add a carrier-level, no-app AI assistant on phone calls in Germany, a major deployment signal for call handling, scheduling, and frontline service workflows. In parallel, 14.ai’s push to replace customer support teams at startups reinforces that AI agents are moving from pilots to staffing substitutes. Cursor’s reported $2B annualized revenue highlights rapid adoption of AI coding tools, nudging AI job risk upward for coding-adjacent tasks (boilerplate, refactors, test generation) while leaving higher-accountability engineering work less exposed. The OpenAI–DoD/Claude switching controversy mainly affects vendor choice, not capability, but it underscores accelerating institutionalization of AI systems and ongoing volatility in tooling ecosystems.
- Referenced news from TechCrunch: Cursor has reportedly surpassed $2B in annualized revenue https://techcrunch.com/2026/03/02/cursor-has-reportedly-surpassed-2b-in-annualized-revenue/
- Referenced news from Wired: What Is That Mysterious Metallic Device US Chief Design Officer Joe Gebbia Is Using? https://www.wired.com/story/joe-gebbia-mystery-metallic-device/
- Referenced news from Wired: This AI Agent Is Ready to Serve, Mid-Phone Call https://www.wired.com/story/deutsche-telekom-elevenlabs-ai-phone-calls-mwc-2026/
- Referenced news from TechCrunch: ChatGPT uninstalls surged by 295% after DoD deal https://techcrunch.com/2026/03/02/chatgpt-uninstalls-surged-by-295-after-dod-deal/
- Referenced news from TechCrunch: No one has a good plan for how AI companies should work with the government https://techcrunch.com/2026/03/02/openai-anthropic-department-of-defense-war-hegseth-ai-companies-work-with-us-government/
- Referenced news from TechCrunch: Users are ditching ChatGPT for Claude. Here’s how to make the switch https://techcrunch.com/2026/03/02/users-are-ditching-chatgpt-for-claude-heres-how-to-make-the-switch/
- Referenced news from MIT Technology Review: OpenAI’s “compromise” with the Pentagon is what Anthropic feared https://www.technologyreview.com/2026/03/02/1133850/openais-compromise-with-the-pentagon-is-what-anthropic-feared/
- Referenced news from TechCrunch: Tech workers urge DOD, Congress to withdraw Anthropic label as a supply chain risk https://techcrunch.com/2026/03/02/tech-workers-urge-dod-congress-to-withdraw-anthropic-label-as-a-supply-chain-risk/
- Referenced news from TechCrunch: A married founder duo’s company, 14.ai, is replacing customer support teams at startups https://techcrunch.com/2026/03/02/a-married-founder-duos-company-14-ai-is-replacing-customer-support-teams-at-startups/
- Referenced news from TechCrunch: Anthropic’s Claude reports widespread outage https://techcrunch.com/2026/03/02/anthropics-claude-reports-widespread-outage/
- Referenced news from MIT Technology Review: I checked out one of the biggest anti-AI protests ever https://www.technologyreview.com/2026/03/02/1133814/i-checked-out-londons-biggest-ever-anti-ai-protest/
- Referenced news from Wired: The Data Centers Have Arrived at the Edge of the Arctic Circle https://www.wired.com/story/ai-supremacy-data-center-expansion-arctic-circle/
- Referenced news from TechCrunch: Google looks to tackle longstanding RCS spam in India — but not alone https://techcrunch.com/2026/03/01/google-looks-to-tackle-longstanding-rcs-spam-in-india-but-not-alone/
- Referenced news from TechCrunch: Investors spill what they aren’t looking for anymore in AI SaaS companies https://techcrunch.com/2026/03/01/investors-spill-what-they-arent-looking-for-anymore-in-ai-saas-companies/
- Referenced news from TechCrunch: OpenAI shares more details about its agreement with the Pentagon https://techcrunch.com/2026/03/01/openai-shares-more-details-about-its-agreement-with-the-pentagon/
- Referenced news from TechCrunch: Anthropic’s Claude rises to No. 1 in the App Store following Pentagon dispute https://techcrunch.com/2026/03/01/anthropics-claude-rises-to-no-2-in-the-app-store-following-pentagon-dispute/
- Referenced news from TechCrunch: SaaS in, SaaS out: Here’s what’s driving the SaaSpocalypse https://techcrunch.com/2026/03/01/saas-in-saas-out-heres-whats-driving-the-saaspocalypse/
- Referenced news from TechCrunch: The trap Anthropic built for itself https://techcrunch.com/2026/02/28/the-trap-anthropic-built-for-itself/
- Referenced news from TechCrunch: Anthropic’s Claude rises to No. 2 in the App Store following Pentagon dispute https://techcrunch.com/2026/02/28/anthropics-claude-rises-to-no-2-in-the-app-store-following-pentagon-dispute/
- Referenced news from TechCrunch: The billion-dollar infrastructure deals powering the AI boom https://techcrunch.com/2026/02/28/billion-dollar-infrastructure-deals-ai-boom-data-centers-openai-oracle-nvidia-microsoft-google-meta/

## Jobs Replaced By AI

- URL: https://ai-job-risk.net/jobs-replaced-by-ai
- What "replaced by AI" means: "Replaced by AI" almost never means an entire profession disappears overnight. What actually happens is task-level: the repeatable, well-defined parts of a job get automated first, head-count growth slows or reverses, entry-level openings thin out, and the remaining work concentrates on judgement, accountability, and dealing with exceptions. We treat a job as "being replaced" when there is documented evidence that its core tasks are being handed to AI at scale, not simply that a model can imitate the output.

### Already being replaced

- Translators and interpreters. (https://ai-job-risk.net/jobs/translator): Neural machine translation turned much of the work from translating-from-scratch into post-editing an AI draft, and the pay structure followed the tasks down. High-volume, lower-stakes content is increasingly handled machine-first with a human checking the output.
- Customer support and call-center agents. (https://ai-job-risk.net/jobs/customer-support-representative): AI chat and voice assistants now resolve a large share of routine, repetitive enquiries end to end, compressing the volume of tickets that ever reach a person and shrinking the case for large first-line teams.
- Data-entry clerks and administrative assistants. (https://ai-job-risk.net/jobs/data-entry-clerk): Document capture, data extraction, and routine back-office processing are exactly the kind of structured, rule-bound work that automation handles well, so the same output now needs fewer hands.
- Copywriters and content writers. (https://ai-job-risk.net/jobs/copywriter): Generative models produce serviceable first drafts of product descriptions, ad copy, and routine marketing content in seconds, pushing demand away from volume writing and toward editing, strategy, and brand judgement.
- Bookkeeping and accounting clerks. (https://ai-job-risk.net/jobs/bookkeeper): Software now reconciles transactions, categorizes entries, and flags anomalies automatically, so a single person supported by automation covers what once took a small team of clerks.
- Proofreaders and copy editors. (https://ai-job-risk.net/jobs/proofreader): AI grammar and editing tools catch a large share of mechanical errors instantly, suppressing demand for routine, first-pass proofreading even where final editorial judgement still matters.
- Telemarketers. (https://ai-job-risk.net/jobs/telemarketer): AI voice agents now handle outbound calling, lead qualification, and routine follow-ups, automating the scripted, high-repetition core of the role.
- Illustrators and graphic designers. (https://ai-job-risk.net/jobs/illustrator): Text-to-image generation produces usable visuals on demand, undercutting commissions for routine, high-volume illustration and template-style design work.
- Paralegals and legal assistants. (https://ai-job-risk.net/jobs/paralegal): Document review, discovery, and first-draft legal research — the high-volume reading work of legal support — are being compressed dramatically by AI systems that scan thousands of pages in hours.

### Being displaced now

- Junior software developers. (https://ai-job-risk.net/jobs/programmer): AI code generation now writes routine implementation, boilerplate, and tests, which hits entry-level programming hardest — the very tasks junior developers used to be hired to do.
- Market research analysts. (https://ai-job-risk.net/jobs/market-research-analyst): Generative AI now drafts research summaries, standardizes reporting, and pulls together comparisons, automating the routine synthesis layer of analyst work.
- Financial analysts and junior bankers. (https://ai-job-risk.net/jobs/financial-analyst): AI tools now assemble pitch materials, build first-pass models, and summarize filings, compressing the hours of routine analysis that junior finance roles were built around.
- Recruiters and HR coordinators. (https://ai-job-risk.net/jobs/recruiter): AI agents now source candidates, screen applications, and schedule interviews, automating large parts of the coordination-heavy core of recruiting.
- SEO specialists and digital marketers. (https://ai-job-risk.net/jobs/seo-specialist): AI-generated answers at the top of search results are absorbing clicks that used to flow to websites, eroding the search-traffic foundation that much SEO and content-marketing work was built on.
- Journalists and news writers. (https://ai-job-risk.net/jobs/journalist): Some outlets now publish AI-assisted or AI-generated articles, while AI search answers pull readers away from news sites, squeezing the advertising economics that fund newsrooms.
- Voice actors and narrators. (https://ai-job-risk.net/jobs/actor): AI voice synthesis now narrates audiobooks, ads, and explainer content, taking on routine narration work that used to be recorded by people.

## Job Pages And Current Scores

### Accountant

- URL: https://ai-job-risk.net/jobs/accountant
- Slug: accountant
- Industry: finance.
- Current score: score 62, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace accountants? Draft journal entries and disclosures get faster, but accounting judgment and control design stay human.
- Editorial overview: Accountants do far more than process transactions. They interpret accounting standards, decide whether treatment matches the economic reality of a transaction, design controls that prevent recurring problems, and explain important accounting issues to management, auditors, and other stakeholders. Their work is about judgment, not just bookkeeping.

### Accounting Clerk

- URL: https://ai-job-risk.net/jobs/accounting-clerk
- Slug: accounting-clerk
- Industry: finance.
- Current score: score 77, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace accounting clerks? Invoice reading, journal entries and bank matching automate first, but final responsibility for money stays human.
- Editorial overview: Accounting clerks do much more than enter vouchers. They process invoices, reimburse expenses, support journal entries, prepare payments, verify documentation, and gather what is needed for the monthly close, all while keeping the company’s money flow recorded in a way that can be explained later. The role involves a large volume of small tasks, but it also carries serious responsibility for numerical consistency and documentation quality.

### Actor

- URL: https://ai-job-risk.net/jobs/actor
- Slug: actor
- Industry: entertainment.
- Current score: score 39, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace actors? Generated faces and voices cut placeholder assets first, while script interpretation and on-set interaction stay human.
- Editorial overview: An actor does far more than read lines from a script. The work involves understanding the intent of the screenplay, syncing with directors and co-stars, and bringing emotion to life within the constraints of a set or stage. Even the same line has to be delivered differently depending on the other performer, camera position, and editing plan, so the job requires judgment that goes far beyond simple recitation.

### Administrative Assistant

- URL: https://ai-job-risk.net/jobs/administrative-assistant
- Slug: administrative-assistant
- Industry: operations.
- Current score: score 87, weekly change +1, week 2026-09-02.
- Latest score explanation: OpenAI’s reported persistent agent work slightly increases automation pressure on calendar management, follow-ups, document prep, and inbox triage. This week’s agent news suggests longer-running office workflows are becoming more deployable, even if governance concerns still slow full replacement.
- Editorial description: Will AI replace administrative assistants? The field fell from 3.5M workers in 2004 to 2.1M in 2024, with medical roles the lone growth pocket.
- Editorial overview: Administrative assistants do much more than create documents or answer phones. They keep work from stalling by handling meeting preparation, calendar coordination, stakeholder communication, paperwork, information gathering, and the prevention of missed steps for managers and departments. Much of the work happens behind the scenes, but it plays a core coordinating role that supports how an organization functions.

### Advertising Specialist

- URL: https://ai-job-risk.net/jobs/advertising-specialist
- Slug: advertising-specialist
- Industry: marketing.
- Current score: score 58, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace advertising specialists? Bid optimization and creative rotation are already automated; message strategy and risk calls are not.
- Editorial overview: Advertising specialists do a great deal more than simply submit assets into media platforms. Their role is to decide which channels to use, what message to run, and when to run it based on the goals of a product or service, then improve the approach based on results. Because the job must balance creative, media characteristics, budget, and legal or brand constraints at the same time, the practical work is highly integrated.

### Agricultural Scientist

- URL: https://ai-job-risk.net/jobs/agricultural-scientist
- Slug: agricultural-scientist
- Industry: agriculture.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace agricultural scientists? Routine analysis speeds up, but rebuilding hypotheses around field reality stays human. See what changes first.
- Editorial overview: Agricultural scientists do much more than analyz crop data. Their work includes designing experiments, reading local field differences, finding causal relationships across many factors, and translating research into methods that producers can actually use.

### AI Engineer

- URL: https://ai-job-risk.net/jobs/ai-engineer
- Slug: ai-engineer
- Industry: technology.
- Current score: score 31, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace AI engineers? Building demos and wrappers is already faster, but evaluation design, guardrails, and cost control stay human work.
- Editorial overview: AI engineers do much more than call a model API. Their role is to decide how a model should be incorporated into a real business problem and what level of accuracy, speed, cost, and safety is realistic. In practice, that means turning ideas into something that can actually run in production, including RAG, agents, evaluation, monitoring, and guardrails.

### Air Traffic Controller

- URL: https://ai-job-risk.net/jobs/air-traffic-controller
- Slug: air-traffic-controller
- Industry: transportation.
- Current score: score 18, weekly change -1, week 2026-09-02.
- Latest score explanation: The OpenAI/Hugging Face hack and broader warnings about AI-driven cybersecurity failures reinforce how unacceptable autonomous errors remain in safety-critical control roles. That slightly lowers near-term replacement risk because air traffic operations require trusted human override, accountability, and resilience under edge cases.
- Editorial description: Will AI replace air traffic controllers? AI handles arrival prediction and route suggestions, but protecting safety margins stays a human call.
- Editorial overview: Air traffic controllers do far more than move aircraft through a queue. They decide the priority of departures, arrivals, and route changes while preserving safety margins across an entire airspace. They continually make safety-side decisions in short time windows while factoring in weather, congestion, equipment conditions, and pilot reports.

### Aircraft Mechanic

- URL: https://ai-job-risk.net/jobs/aircraft-mechanic
- Slug: aircraft-mechanic
- Industry: transportation.
- Current score: score 22, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace aircraft mechanics? Fault-code triage and log comparison automate, but the call on whether a plane flies stays with a person.
- Editorial overview: Aircraft mechanics do much more than replace parts according to maintenance manuals. They create airworthy conditions by judging the state of the aircraft, operating conditions, and warning signs, while tying inspection, diagnosis, maintenance records, and return-to-service decisions together.

### Animator

- URL: https://ai-job-risk.net/jobs/animator
- Slug: animator
- Industry: creative.
- Current score: score 73, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace animators? Interpolation and cleanup are being absorbed; timing, performance design and dramatic intent are not. See where the line sits.
- Editorial overview: An animator is not simply someone who draws a lot of frames. The role is about reading emotion, weight, timing, and the function of a shot, then turning that into motion that works. The responsibility lies less in output volume than in judging what kind of movement will actually communicate something.

### Anthropologist

- URL: https://ai-job-risk.net/jobs/anthropologist
- Slug: anthropologist
- Industry: science.
- Current score: score 27, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace anthropologists? AI can organize field notes, but observation, ethical judgment, and reading what went unsaid remain human work.
- Editorial overview: Anthropologists do a great deal more than studie culture in the abstract. They work to understand the background behind the customs, values, and behavior of communities and groups through on-site observation and interviews. Their role is to capture context that numbers alone cannot reveal and carefully interpret ways of life that are easy to misunderstand from the outside.

### Architect

- URL: https://ai-job-risk.net/jobs/architect
- Slug: architect
- Industry: construction.
- Current score: score 36, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace architects? Draft floor plans and code checks get faster, but resolving budget, site and buildability conflicts stays human work.
- Editorial overview: Architects do much more than draw plans. They design buildings that can actually work in the real world by balancing site conditions, regulations, structure, building systems, construction cost, and the client's goals at the same time. Their responsibility extends beyond visual proposals to permit applications, detailed design, and construction supervision, making space viable in a social and practical sense.

### Archivist

- URL: https://ai-job-risk.net/jobs/archivist
- Slug: archivist
- Industry: education.
- Current score: score 43, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace archivists? OCR, identifiers and summaries automate well; provenance, preservation units and access rules still need judgment.
- Editorial overview: Archivists do much more than preserv documents and records. They organize materials so they can be referenced in the future, manage the context in which those materials were created, and decide how they should be used. They are less like storage staff and more like professionals responsible for preserving the reliability and long-term usability of records.

### Astronomer

- URL: https://ai-job-risk.net/jobs/astronomer
- Slug: astronomer
- Industry: science.
- Current score: score 22, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace astronomers? Large-scale data screening gets faster, but deciding what a signal means under uncertainty stays human. See what changes.
- Editorial overview: Astronomers do far more than collect and classify observational data. Their role is to connect observation, theory, and simulation in order to understand what is happening in the universe and what assumptions are justified. They work with limited signals and incomplete data, so deciding how far an interpretation can go is part of the profession.

### Athletic Coach

- URL: https://ai-job-risk.net/jobs/athletic-coach
- Slug: athletic-coach
- Industry: hospitality.
- Current score: score 16, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace sports coaches? Video comparison and opponent analysis automate first; turning that into advice for one athlete does not.
- Editorial overview: Sports coaches do a great deal more than assign practice plans. They support each athlete's technical, physical, and psychological development by observing condition and deciding what kind of training will matter for whom. The same drill can affect different athletes in very different ways.

### Auditor

- URL: https://ai-job-risk.net/jobs/auditor
- Slug: auditor
- Industry: finance.
- Current score: score 49, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace auditors? Sample selection and prior-period comparisons go first, while setting key issues and weighing evidence stay human calls.
- Editorial overview: Auditors do a great deal more than compare documents. They set the important issues to investigate, weigh the quality and consistency of evidence, evaluate whether internal controls actually work in practice, and negotiate corrective action with the relevant teams. Their role involves more than counting problems; it also involves deciding which ones matter and what should happen next.

### Automotive Technician

- URL: https://ai-job-risk.net/jobs/automotive-technician
- Slug: automotive-technician
- Industry: manufacturing.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace automotive technicians? Fault codes and estimate drafts speed up, but cause identification on the actual vehicle stays human.
- Editorial overview: Automotive technicians do far more than pas vehicles through inspection. Their job is to identify the cause of problems by watching for unusual noises, vibration, warning lights, driving feel, and electronic-control behavior, then restore the vehicle to a condition where it can be driven safely. They are responsible not only for the repair work itself, but also for explaining necessary repairs to the owner.

### Baker

- URL: https://ai-job-risk.net/jobs/baker
- Slug: baker
- Industry: hospitality.
- Current score: score 36, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace bakers? Weighing and logging automate well, but fermentation judgment and baking adjustment still run on touch, smell and timing.
- Editorial overview: Bakers do far more than follow a recipe. They create final quality by watching dough condition, fermentation, temperature, humidity, and crust color. Even with the same formula, the state of the dough changes day by day, so value comes from accumulated sensory judgment and process control.

### Bank Teller

- URL: https://ai-job-risk.net/jobs/bank-teller
- Slug: bank-teller
- Industry: finance.
- Current score: score 69, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace bank tellers? ATMs and apps already absorb deposits and transfers, leaving identity checks and suspicious activity to people.
- Editorial overview: Bank tellers do far more than accept deposits and withdrawals. They also verify identity, explain procedures, handle customer questions, process forms, and serve as the first point of judgment when something about a transaction feels unusual. The work blends routine financial processing with face-to-face risk awareness.

### Barista

- URL: https://ai-job-risk.net/jobs/barista
- Slug: barista
- Industry: hospitality.
- Current score: score 41, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace baristas? Recipe management and stock forecasting automate first; taste adjustment and the room's atmosphere stay human work.
- Editorial overview: Baristas do much more than make drinks. They create the full café experience by adjusting to bean condition, extraction, timing of service, and the right distance in conversation. Both flavor and human interaction are evaluated at the same time.

### Bartender

- URL: https://ai-job-risk.net/jobs/bartender
- Slug: bartender
- Industry: hospitality.
- Current score: score 26, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace bartenders? Recipe management and stock forecasting automate; reading the guest and the mood of the room is what remains.
- Editorial overview: Bartenders do far more than make drinks. They create both a glass and an atmosphere by reading a guest's mood, the flow of conversation, and the temperature of the room. Much of the value lies not in the drink itself, but in how it is served and how the space is shaped around it.

### Biologist

- URL: https://ai-job-risk.net/jobs/biologist
- Slug: biologist
- Industry: science.
- Current score: score 30, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace biologists? Image classification and sequence work speed up, but experiment design and reproducibility stay human. See what shifts first.
- Editorial overview: Biologists do much more than collect samples or run assays. Their role is to understand living systems by connecting observations, experiments, and environmental or cellular context. Because living systems are variable and often sensitive to small condition changes, the job depends heavily on judgment about what differences matter.

### Bookkeeper

- URL: https://ai-job-risk.net/jobs/bookkeeper
- Slug: bookkeeper
- Industry: finance.
- Current score: score 78, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace bookkeepers? Software drafts entries and reads receipts, but evidence-to-ledger checks and exception review still need a person.
- Editorial overview: Bookkeepers do much more than post daily entries. They also keep records aligned with supporting evidence, monitor balances, prevent duplicate or missing entries, and keep the books building correctly toward each close. The role is less about speed alone and more about preserving the integrity of the ledger over time.

### Brand Manager

- URL: https://ai-job-risk.net/jobs/brand-manager
- Slug: brand-manager
- Industry: marketing.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace brand managers? Taglines and competitor summaries now generate in bulk, but deciding what the brand promises stays a human call.
- Editorial overview: Brand managers do much more than supervise advertising language. Their role is to define how the company should be positioned in the market, what promises it should make to customers, and what kind of experience should reinforce trust over time. The work sits at the center of product planning, sales, customer support, PR, and marketing, and it is fundamentally about making consistent cross-functional decisions.

### Bus Driver

- URL: https://ai-job-risk.net/jobs/bus-driver
- Slug: bus-driver
- Industry: transportation.
- Current score: score 61, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace bus drivers? Lane keeping and route optimization are automated; boarding decisions and in-cabin incidents remain human work.
- Editorial overview: Bus drivers do much more than operate a vehicle. They keep service stable by getting passengers on and off safely and by adjusting operation to road conditions. Timeliness, safety, passenger handling, and checks inside and outside the vehicle all happen at once.

### Business Analyst

- URL: https://ai-job-risk.net/jobs/business-analyst
- Slug: business-analyst
- Industry: consulting.
- Current score: score 73, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace business analysts? Gartner projects 40 percent of enterprise applications will carry AI agents by the end of 2026. See what work survives.
- Editorial overview: Business analysts do a great deal more than compile documents. They organize operational issues, workflows, data, and stakeholder interests to define what the real problem is and what requirements would actually lead to improvement. Their responsibility is not analysis for its own sake, but turning ambiguity into questions that decision-makers can use.

### Call Center Agent

- URL: https://ai-job-risk.net/jobs/call-center-agent
- Slug: call-center-agent
- Industry: marketing.
- Current score: score 90, weekly change +1, week 2026-09-02.
- Latest score explanation: Persistent AI agents and easier local chatbot deployment both strengthen automation of scripted customer interactions, knowledge retrieval, and routine escalation handling. This week’s developments point to broader deployment readiness for always-on support workflows, nudging risk up relative to other jobs.
- Editorial description: Will AI replace call center agents? Klarna said its AI did the work of 700 staff, then quietly rehired humans. See which calls still need one.
- Editorial overview: Call center agents do far more than answer questions over the phone. Within a limited amount of call time, they have to grasp the caller's emotions, urgency, and situation, provide the right guidance and confirmations, and move the caller toward the next action. Because they work with signals such as tone of voice, pauses, and signs of panic that text doesn't reveal, the job involves a great deal of phone-specific judgment.

### Career Counselor

- URL: https://ai-job-risk.net/jobs/career-counselor
- Slug: career-counselor
- Industry: education.
- Current score: score 28, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace career counselors? Resume drafts and job comparisons come fast; clarifying what a client actually values and fears does not.
- Editorial overview: Career counselors do far more than recommend job openings. They help organize a client’s experience, values, anxieties, strengths, and real-life constraints, then think through career paths and work styles together with them. Their role includes decision support, application preparation, interview preparation, and emotional guidance, not just information sharing.

### Carpenter

- URL: https://ai-job-risk.net/jobs/carpenter
- Slug: carpenter
- Industry: construction.
- Current score: score 24, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace carpenters? Material takeoffs and cutting instructions generate fast, but fitting a site that never matches the drawing does not.
- Editorial overview: Carpenters do much more than cut and assemble wood. They read drawings, think about structure, safety, fit, appearance, and coordination with other trades, and then adjust dimensions on site to make the space work. Their role includes not only new construction, but also renovations, repairs, and on-site custom fitting.

### Chef

- URL: https://ai-job-risk.net/jobs/chef
- Slug: chef
- Industry: hospitality.
- Current score: score 23, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace chefs? Food-cost and popularity analysis automate, but flavor direction and holding a kitchen together do not. See which tasks go first.
- Editorial overview: Chefs do much more than cook dishes. They design the dining experience by weighing ingredient condition, kitchen flow, timing of service, and the overall direction of the restaurant. Flavor creation and kitchen leadership are inseparable in the role.

### Chemist

- URL: https://ai-job-risk.net/jobs/chemist
- Slug: chemist
- Industry: science.
- Current score: score 35, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace chemists? AI proposes candidate reactions, but safety, reproducibility, and scale-up judgment decide what works in the plant.
- Editorial overview: Chemists do far more than run reactions. They decide which conditions are chemically plausible, safe, and reproducible, while linking measurements back to material behavior and process constraints. The work often spans both the laboratory and the realities of eventual production.

### Civil Drafter

- URL: https://ai-job-risk.net/jobs/civil-drafter
- Slug: civil-drafter
- Industry: construction.
- Current score: score 73, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace civil drafters? Layers, sections and repeat revisions automate first, but reading design intent into usable drawings does not.
- Editorial overview: Civil drafters do far more than produc clean-looking drawings. Their job is to read the engineer's intent and translate roads, bridges, earthworks, drainage, and structural dimensions into drawings that can actually be used in construction. Their role supports quality not only through linework, but through consistency across drawings, notes, scales, quantities, and overall readability.

### Civil Engineer

- URL: https://ai-job-risk.net/jobs/civil-engineer
- Slug: civil-engineer
- Industry: construction.
- Current score: score 32, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace civil engineers? Analysis support, comparisons and documentation speed up; weighing maintenance and public duty does not.
- Editorial overview: Civil engineers do much more than calculate structures or draw infrastructure layouts. Their work is to design roads, bridges, embankments, drainage systems, and other public works in a way that balances safety, long-term use, maintenance, cost, the surrounding environment, and social impact. They are responsible not only for whether something can be built, but for whether it will continue to function well over time.

### Claims Adjuster

- URL: https://ai-job-risk.net/jobs/claims-adjuster
- Slug: claims-adjuster
- Industry: finance.
- Current score: score 59, weekly change +2, week 2026-09-02.
- Latest score explanation: A WIRED report focused directly on claims adjusters showed overwhelmingly negative worker reactions to AI and implied active use of AI in core claims review workflows. Because this job includes document assessment, case triage, and standardized decisions that insurers are visibly trying to automate, the score moves up modestly.
- Editorial description: Will AI replace claims adjusters? Document sorting and baseline payout estimates automate; fact-finding and policy boundary calls still need a person.
- Editorial overview: Claims adjusters do a great deal more than calculate insurance payouts. Their job is to confirm the facts of an accident or loss, determine whether payment should be made and whether the amount is appropriate under the policy terms, and explain that decision to related parties when necessary. The role protects the fairness of payments by covering claim review, organization of field information, detection of suspicious patterns, and customer communication.

### Climate Analyst

- URL: https://ai-job-risk.net/jobs/climate-analyst
- Slug: climate-analyst
- Industry: environment.
- Current score: score 55, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace climate analysts? Scenario aggregation is getting fast, but deciding which climate risks actually move a business stays human.
- Editorial overview: Climate analysts do far more than track long-term changes in temperature and rainfall. Their job is to analyze what those changes mean for companies, local governments, infrastructure, and finance in terms of both risk and opportunity. The role requires both scientific literacy and the ability to translate that science into business and policy decisions.

### Cloud Engineer

- URL: https://ai-job-risk.net/jobs/cloud-engineer
- Slug: cloud-engineer
- Industry: technology.
- Current score: score 38, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace cloud engineers? Standard templates and first-pass triage automate; redundancy, security boundaries and incident calls do not.
- Editorial overview: Cloud engineers do a great deal more than spin up servers. Their job is to design cloud infrastructure that keeps applications available, secure, and financially sustainable in operation. That means thinking in an integrated way about networking, permissions, monitoring, backups, availability, and disaster recovery, not just performing infrastructure setup tasks.

### Compensation Analyst

- URL: https://ai-job-risk.net/jobs/compensation-analyst
- Slug: compensation-analyst
- Industry: consulting.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace compensation analysts? Market-range comparisons and raise simulations automate; setting pay policy lines still needs people.
- Editorial overview: A compensation analyst does far more than tabulate pay data. The role is about designing pay levels that protect both hiring competitiveness and internal consistency while balancing market benchmarks, internal grades, role responsibility, fairness, and labor-cost constraints. More than calculating numbers, it is a role with real responsibility for drawing the line.

### Composer

- URL: https://ai-job-risk.net/jobs/composer
- Slug: composer
- Industry: entertainment.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace composers? Generic BGM, drafts and loop work automate fast, while emotional design and project-specific sound identity stay human.
- Editorial overview: A composer does far more than write melodies. The job is to design how emotion should move across a work, translate vague requests from directors or clients into sound, and build music with arrangement, recording, and mix in mind from the start.

### Construction Manager

- URL: https://ai-job-risk.net/jobs/construction-manager
- Slug: construction-manager
- Industry: construction.
- Current score: score 32, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace construction managers? Schedules, daily reports, and safety drafts automate; deciding priorities when a site drifts does not.
- Editorial overview: Construction managers do a great deal more than supervis sites. Their work is to make decisions that keep a project moving as planned while balancing schedule, quality, safety, cost, subcontractors, and client demands at the same time. They are responsible both for building according to the drawings and for absorbing delays and mismatches that arise on site and keeping the whole project moving forward.

### Construction Worker

- URL: https://ai-job-risk.net/jobs/construction-worker
- Slug: construction-worker
- Industry: construction.
- Current score: score 45, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace construction workers? Delivery records and daily reports go first; moving safely through a site that changes daily does not.
- Editorial overview: Construction workers do a great deal more than simply move their bodies on site. They help form the operational base that keeps a site running through material handling, scaffolding, cleanup, trade support, safety attention, and responses to changes in sequencing. Even while taking direction from foremen and specialists, their ability to notice danger and inefficiency in the moment matters greatly.

### Content Writer

- URL: https://ai-job-risk.net/jobs/content-writer
- Slug: content-writer
- Industry: creative.
- Current score: score 81, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace content editors? Drafts, headlines and summaries are fast now, but search intent, structure and fact-checking need an editor.
- Editorial overview: Content editors do much more than simply write articles. Their job is to decide what should become a heading, what should be explored in depth, and what tone should be used, based on search intent and the editorial policy of the publication. Their value includes line editing, restructuring, fact-checking, title adjustments, publication decisions, and post-publication improvement.

### Cook

- URL: https://ai-job-risk.net/jobs/cook
- Slug: cook
- Industry: hospitality.
- Current score: score 39, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace cooks? Order forecasting and standardized recipe control automate, but finishing judgment, sanitation and kitchen flow stay with people.
- Editorial overview: A cook does more than follow recipes. The job is to finish dishes while balancing hygiene and quality, paying close attention to the condition of the ingredients and the timing of service. In the flow of a kitchen, cooks are constantly judging what must be rushed and what must be handled with extra care.

### Copywriter

- URL: https://ai-job-risk.net/jobs/copywriter
- Slug: copywriter
- Industry: marketing.
- Current score: score 84, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace copywriters? AI mass-produces headlines and CTA options, so value moves to deciding the core message and how far a brand goes.
- Editorial overview: Copywriters are not simply people who write well. They design what impression should be created and how it should be created with a small number of words. Through ad headlines, landing-page messaging, brand slogans, product names, banner copy, and email subject lines, they influence emotions and action. Even though the word count is small, the impact on business results can be large.

### Court Reporter

- URL: https://ai-job-risk.net/jobs/court-reporter
- Slug: court-reporter
- Industry: legal.
- Current score: score 77, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace court reporters? Draft transcription automates, but verifying a record that must hold up procedurally stays human. See the split.
- Editorial overview: A court reporter handles more than raw transcription. The job is to record who said what, in what order, and what must remain as an official record, while distinguishing speakers, confirming specialized terms, and fitting speech into formal court-record formats. The role supports the foundation of judicial procedure.

### Curriculum Developer

- URL: https://ai-job-risk.net/jobs/curriculum-developer
- Slug: curriculum-developer
- Industry: education.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace curriculum developers? Syllabus drafts and sample assessments come fast; learning sequence and assessment validity do not.
- Editorial overview: Curriculum developers do far more than arrange teaching materials. They define learning goals, determine the order in which people should learn, decide how mastery will be assessed, and design combinations of materials and activities that actually lead to learning outcomes. Whether the setting is school, corporate training, or certification programs, they shape the structure of the learning experience.

### Customer Success Manager

- URL: https://ai-job-risk.net/jobs/customer-success-manager
- Slug: customer-success-manager
- Industry: marketing.
- Current score: score 28, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace customer success managers? Onboarding drafts and health-score alerts automate; diagnosing stalled adoption still needs a person.
- Editorial overview: A customer success manager is not simply an extension of support. The role is to ensure that customers after signing can fully adopt the product or service, achieve their goals, and reach a state that leads to renewals or expansion. Because the work involves onboarding support, usage recommendations, checking product data, renewal discussions, and internal feedback, it is better understood as an upstream, partnership-based role rather than simple support.

### Customer Support

- URL: https://ai-job-risk.net/jobs/customer-support
- Slug: customer-support
- Industry: marketing.
- Current score: score 86, weekly change +1, week 2026-09-02.
- Latest score explanation: This week’s persistent-agent reporting raises the likelihood that AI systems can handle multi-step support actions rather than just answer single questions. That matters for customer support work involving ticket updates, follow-up messages, and account workflow handling, so risk edges higher.
- Editorial description: Will AI replace customer support? Chatbots handle hours, pricing, and basic how-to questions; the stuck and angry cases still reach a human.
- Editorial overview: Customer support goes beyond answering inquiries. It is about identifying where users are stuck, where trust has broken down, and what kind of explanation will help them feel confident enough to move forward. The job includes handling inquiries, requesting internal investigation, coordinating with other teams, and organizing measures to prevent recurrence. In practice, it is closer to a role focused on restoring trust through problem resolution.

### Customer Support Representative

- URL: https://ai-job-risk.net/jobs/customer-support-representative
- Slug: customer-support-representative
- Industry: marketing.
- Current score: score 90, weekly change +1, week 2026-09-02.
- Latest score explanation: OpenAI’s agent progress and widespread discussion of enterprise agent orchestration support more automation of repetitive support conversations, troubleshooting scripts, and handoff decisions. The job’s task mix is directly exposed to these deployment signals, so its score increases slightly.
- Editorial description: Will AI replace customer support reps? Bots absorb the self-service FAQs, so the cases reaching people get harder and more emotionally charged.
- Editorial overview: Customer support representatives work on the front line of inquiry handling, dealing with individual cases one by one. Their job involves more than replying according to a manual. They have to read urgency and emotional tone from the customer’s wording and manner, ask for the information that is needed, and connect the case to the right team or next step. In practice, they are often expected to balance both processing volume and customer satisfaction, which makes judgment especially important.

### Cybersecurity Analyst

- URL: https://ai-job-risk.net/jobs/cybersecurity-analyst
- Slug: cybersecurity-analyst
- Industry: technology.
- Current score: score 24, weekly change -1, week 2026-09-02.
- Latest score explanation: Warnings of an AI-driven cybersecurity crisis and reports of agents hacking systems increase demand for human defenders rather than reducing it. While AI assists detection and response, this week’s news emphasizes oversight, investigation, and adversarial judgment that still keep analysts in the loop.
- Editorial description: Will AI replace cybersecurity analysts? Alert, CVE and IOC matching automate, but false-positive calls and containment scope stay human.
- Editorial overview: Cybersecurity analysts do far more than read alerts. Their job is to look at logs, vulnerabilities, permissions, communications, and usage patterns and judge where the real risk lies and what should be protected first. What they protect includes not only systems, but also the business, customer information, and business continuity.

### Data Analyst

- URL: https://ai-job-risk.net/jobs/data-analyst
- Slug: data-analyst
- Industry: technology.
- Current score: score 84, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace data analysts? Aggregation, dashboards and SQL drafting automate, but defining which number actually matters for a decision does not.
- Editorial overview: Data analysts do far more than arrange numbers neatly. They translate changes in data into forms that decision-makers can actually use. Looking at metrics such as sales, retention, churn, inquiries, inventory, and advertising spend, they provide material for judging what is happening and where attention should go first.

### Data Entry Clerk

- URL: https://ai-job-risk.net/jobs/data-entry-clerk
- Slug: data-entry-clerk
- Industry: technology.
- Current score: score 82, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace data entry clerks? Tools like ABBYY and UiPath claim 90 to 99 percent accuracy on standard forms. See what still needs a person.
- Editorial overview: Data entry clerks do a great deal more than type information into a system. In practice, they also handle transcription from forms, importing CSV and list data, checking for missing fields, spotting inconsistent notation, and preparing information in a form downstream processes can actually use. By 2026, intelligent document processing platforms rated by Gartner, including ABBYY, Hyperscience, and UiPath, handle the standardized version of that work at scale.

### Data Scientist

- URL: https://ai-job-risk.net/jobs/data-scientist
- Slug: data-scientist
- Industry: technology.
- Current score: score 37, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace data scientists? AutoML commoditizes model building, so problem setting and evaluation design become the real differentiator.
- Editorial overview: Being a data scientist goes beyond building machine-learning models. In practice, the role is to determine which kinds of prediction or optimization create business value, confirm what usable data exists, choose evaluation metrics, and design a solution that can stand up in live operation. Beyond math and implementation, deciding what should be solved is a central part of the job.

### Database Administrator

- URL: https://ai-job-risk.net/jobs/database-administrator
- Slug: database-administrator
- Industry: technology.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace database administrators? Index and SQL tuning ideas are easy to generate; judging what breaks writes or recovery is not.
- Editorial overview: Database administrators do much more than manage database servers. Their role is to keep business-critical information safe and usable by protecting data integrity, availability, performance, backups, permissions, and recovery procedures. Because the impact of database failures is often large, simply staying quiet and stable is itself a major outcome in this role.

### Delivery Driver

- URL: https://ai-job-risk.net/jobs/delivery-driver
- Slug: delivery-driver
- Industry: logistics.
- Current score: score 73, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace delivery drivers? Routing and parcel-volume forecasting are already algorithmic; handoff rules and road hazards still need a driver.
- Editorial overview: Delivery drivers do a great deal more than move parcels from one place to another. They complete deliveries safely and on time while accounting for traffic, load conditions, destination-specific requirements, timed delivery windows, absence cases, and safe driving. Their responsibility is less about driving itself than about making sure delivery still works when exceptions happen.

### Dentist

- URL: https://ai-job-risk.net/jobs/dentist
- Slug: dentist
- Industry: healthcare.
- Current score: score 15, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace dentists? Image review and record organization speed up, but treatment boundaries and patient buy-in stay with the dentist.
- Editorial overview: Dentists do far more than examine the mouth and decide on a treatment plan. They also have to explain treatment in a way patients can stick with, and carry out procedures while managing pain and anxiety. Even as support for image reading and record-keeping increases, the role of weighing symptoms, lifestyle habits, and overall health within a limited appointment window is not easy to replace.

### Detective

- URL: https://ai-job-risk.net/jobs/detective
- Slug: detective
- Industry: legal.
- Current score: score 21, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace detectives? Footage review and record search automate first, but deciding what to verify next in the field stays human work.
- Editorial overview: A detective does more than gather facts. The job is about advancing an investigation while judging which information has evidentiary meaning and what is still missing, combining desk research, field observation, witness interviews, and legal procedure. The profession stands on the integration of those elements, not on data collection alone.

### DevOps Engineer

- URL: https://ai-job-risk.net/jobs/devops-engineer
- Slug: devops-engineer
- Industry: technology.
- Current score: score 39, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace DevOps engineers? Pipeline and script first drafts are quick, but deciding how much to automate and where review stays is not.
- Editorial overview: DevOps engineers do far more than introduce tools. Their job is to build systems that let teams move from development to release, monitoring, and incident response quickly and safely. They connect CI/CD, environment drift, rollback, observability, and developer experience while managing both change speed and accident rate across the team.

### Digital Marketer

- URL: https://ai-job-risk.net/jobs/digital-marketer
- Slug: digital-marketer
- Industry: marketing.
- Current score: score 77, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace digital marketers? Ad copy, budget adjustment and reporting automate, but connecting channels into one funnel does not. See what remains.
- Editorial overview: Digital marketers do much more than run web ads. Their role is to improve the entire customer journey, from acquisition to ongoing use, by connecting ads, landing pages, email, CRM, in-app flows, and measurement systems. The core of the work is not simply watching channel-level numbers, but identifying where prospects are dropping off across the full funnel.

### Diplomat

- URL: https://ai-job-risk.net/jobs/diplomat
- Slug: diplomat
- Industry: government.
- Current score: score 14, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace diplomats? Translation and briefing drafts get faster, but judging what to say, to whom, and when stays a human decision.
- Editorial overview: Diplomats are not simply people who can converse in foreign languages. Their job is to understand another country’s institutions, politics, culture, and negotiating context, then reconcile interests without damaging their own country’s position. They gather information, analyze it, write documents, negotiate, and respond to crises while reading both what counterparts truly mean and what they are saying publicly.

### Doctor

- URL: https://ai-job-risk.net/jobs/doctor
- Slug: doctor
- Industry: healthcare.
- Current score: score 20, weekly change +1, week 2026-09-02.
- Latest score explanation: A new paper highlighted this week argues AI can outperform doctors on some diagnostic and clinical reasoning tasks, increasing automation pressure on information-heavy parts of care. The score rises only slightly because liability, patient trust, and hands-on medical responsibility still limit near-term replacement.
- Editorial description: Will AI replace doctors? Differentials, chart summaries and referral drafts get faster; urgency judgment and treatment decisions stay with you.
- Editorial overview: Doctors do much more than identify diseases. Their work is to weigh symptoms, test results, medical history, living conditions, and urgency together in order to decide what to prioritize now, what to rule out, and how far to intervene. Their responsibility extends beyond diagnosis to explanation, treatment planning, follow-up, and coordination with other departments.

### Economist

- URL: https://ai-job-risk.net/jobs/economist
- Slug: economist
- Industry: finance.
- Current score: score 38, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace economists? CPI, GDP and jobs summaries are already automated; explaining why markets reacted differently is still human work.
- Editorial overview: Economists read the macro environment, growth, inflation, interest rates, employment, and exchange rates, and translate it into forms that businesses and investors can use to make decisions. They do more than introduce statistics. They connect multiple indicators and explain how current shifts are likely to affect markets and business outcomes.

### Editor

- URL: https://ai-job-risk.net/jobs/editor
- Slug: editor
- Industry: media.
- Current score: score 62, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace editors? Outlines, headlines and rewrite suggestions automate, but deciding what to cut and what is safe to publish does not.
- Editorial overview: Editors are not simply people who correct manuscripts. They stand between an idea and its readers and shape information into something genuinely valuable. They are responsible for quality through choices about framing, structure, depth of argument, heading design, fact-checking, and publication standards. In many cases, they work further upstream than writers and help set the quality level of the publication itself.

### Electrical Engineer

- URL: https://ai-job-risk.net/jobs/electrical-engineer
- Slug: electrical-engineer
- Industry: technology.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace electrical engineers? Component search and standard circuit drafts automate; heat, noise, and field constraints need judgment.
- Editorial overview: Electrical engineers do much more than draw circuits and equipment layouts. They design systems that must satisfy safety, maintainability, cost, delivery time, and installation conditions at the same time. Because they judge voltage, current, heat, noise, component life, and regulatory compliance together, their value lies less in automatic design alone and more in reconciling field constraints.

### Electrician

- URL: https://ai-job-risk.net/jobs/electrician
- Slug: electrician
- Industry: construction.
- Current score: score 11, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace electricians? McKinsey projects a gap of roughly 130,000 electricians by 2030, driven by the data-center construction boom.
- Editorial overview: Electricians do a great deal more than connect wires. They work from drawings, site conditions, safety rules, and legal standards to decide where wiring should run, where circuits should branch, and how power should be supplied to each piece of equipment. By 2026 the trade sits at the center of the AI data-center construction boom itself, with McKinsey projecting a gap of roughly 130,000 electricians by 2030 as companies like Amazon, Meta, and Microsoft race to build capacity.

### Elevator Technician

- URL: https://ai-job-risk.net/jobs/elevator-technician
- Slug: elevator-technician
- Industry: construction.
- Current score: score 21, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace elevator technicians? Remote monitoring sorts alarms and parts lists, but the call to stop or restart a unit stays on site.
- Editorial overview: Elevator technicians do much more than repair lifts. They maintain machines, controls, safety devices, building conditions, and user safety in a way that keeps critical equipment operating safely and restores it reliably when problems occur. The role carries serious responsibility across statutory inspections, parts replacement, breakdown response, and modernization decisions.

### Energy Engineer

- URL: https://ai-job-risk.net/jobs/energy-engineer
- Slug: energy-engineer
- Industry: energy.
- Current score: score 30, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace energy engineers? Routine simulations and comparison studies automate, but design under site constraints and regulation stays human.
- Editorial overview: Energy engineers design systems that connect power generation, storage, heat utilization, equipment efficiency, and demand control so energy can be used safely and without waste. Their job involves more than picking equipment; it also involves deciding which approach is realistic while balancing cost, regulation, supply stability, and on-site constraints.

### Environmental Scientist

- URL: https://ai-job-risk.net/jobs/environmental-scientist
- Slug: environmental-scientist
- Industry: environment.
- Current score: score 30, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace environmental scientists? Document search and charts speed up, but choosing sites and judging local regulation stays human.
- Editorial overview: Environmental scientists do much more than collect measurements about air, water, soil, and ecosystems. Their job is to evaluate how those changes affect human life and business activity by combining measurement, field observation, regulation, and reporting into practical risk judgment.

### Farmer

- URL: https://ai-job-risk.net/jobs/farmer
- Slug: farmer
- Industry: agriculture.
- Current score: score 50, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace farmers? Irrigation, monitoring and forecasting automate first, but deciding when conditions call for a different step does not.
- Editorial overview: A farmer does more than repeat fixed work in the field. The job involves reading crop condition, weather shifts, timing, labor limits, and sales priorities, then changing what to do and when to do it. Farming is not only production work, but also operational judgment and business judgment at the same time.

### Fashion Designer

- URL: https://ai-job-risk.net/jobs/fashion-designer
- Slug: fashion-designer
- Industry: creative.
- Current score: score 45, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace fashion designers? Mood boards, colorways and trend summaries generate fast; sizing, material cost and factory limits do not.
- Editorial overview: A fashion designer does more than imagine attractive clothes. The role is about balancing a brand's world, the wearer's real life, material characteristics, price range, and production conditions so that what looks appealing can also be sold and made. The job carries responsibility both for ideas and for deciding which ideas should survive into real products.

### Film Director

- URL: https://ai-job-risk.net/jobs/film-director
- Slug: film-director
- Industry: entertainment.
- Current score: score 23, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace film directors? Rough visual generation is automating, but script interpretation, set priorities and the final form are not.
- Editorial overview: A film director does much more than decide what to shoot in front of a camera. The job is to define how a script should be interpreted and then unify decisions across acting, cinematography, art direction, editing, and music into a single work. Directors are also responsible for deciding what to prioritize within limited time and budgets while holding the project’s overall direction together.

### Financial Analyst

- URL: https://ai-job-risk.net/jobs/financial-analyst
- Slug: financial-analyst
- Industry: finance.
- Current score: score 67, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace financial analysts? Report production and variance extraction automate first; testing assumptions and framing decisions do not.
- Editorial overview: Financial analysts do much more than summarize numbers. They break down the drivers behind financial results, test assumptions, connect accounting figures to business KPIs, and turn financial information into issues and recommendations that management can actually use.

### Firefighter

- URL: https://ai-job-risk.net/jobs/firefighter
- Slug: firefighter
- Industry: public-service.
- Current score: score 10, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace firefighters? BLS projects 3% growth and about 27,100 openings a year through 2034. See what AI detection tools actually change.
- Editorial overview: Firefighters do much more than put out fires. They make life-or-death decisions under severe time pressure across fires, rescues, emergency medical calls, disaster response, and scene safety. The moment they arrive, they have to read the situation, set priorities, and decide who should move where.

### Fisherman

- URL: https://ai-job-risk.net/jobs/fisherman
- Slug: fisherman
- Industry: agriculture.
- Current score: score 42, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace fishermen? Fish-finding, routing and record support keep improving, but sea conditions, crew safety and sale timing stay human.
- Editorial overview: A fisherman does more than locate fish and bring them back. The work includes reading the sea, judging safety, managing the vessel and crew, and making decisions that connect catch, condition, and sale. It is a job where field judgment matters at every stage.

### Fitness Trainer

- URL: https://ai-job-risk.net/jobs/fitness-trainer
- Slug: fitness-trainer
- Industry: hospitality.
- Current score: score 20, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace fitness trainers? Standard programs and data logging automate, but form correction and load judgment need eyes on the client.
- Editorial overview: A fitness trainer does more than hand out workout plans. The job is to guide people safely toward results while watching their physical capacity, posture, willingness to continue, and day-to-day condition. Exercise programming and ongoing support are tightly linked in this role.

### Flight Attendant

- URL: https://ai-job-risk.net/jobs/flight-attendant
- Slug: flight-attendant
- Industry: transportation.
- Current score: score 26, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace flight attendants? Multilingual guidance and seat information automate, but cabin safety and emergency response stay with crew.
- Editorial overview: Flight attendants are both customer-facing service staff and cabin safety personnel. Through boarding guidance, awareness of passenger condition, emergency evacuation, and onboard service, they balance safety and comfort in a tightly constrained environment.

### Game Developer

- URL: https://ai-job-risk.net/jobs/game-developer
- Slug: game-developer
- Industry: technology.
- Current score: score 59, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace game developers? Tool scripts, rough UI and debug support get faster; making difficulty, pacing and fun land stays human work.
- Editorial overview: Game developers are more than programmers. Their work is to make a game feel fun through game feel, pacing, difficulty balance, reward design, presentation, and optimization. Unlike standard business software, it is not enough for a game to simply work correctly. Quality is determined by whether it feels good the moment someone touches it and whether it gives people a reason to keep playing.

### Geologist

- URL: https://ai-job-risk.net/jobs/geologist
- Slug: geologist
- Industry: science.
- Current score: score 26, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace geologists? Mapping and report search speed up, but reading outcrops with limited data and strong local variation does not.
- Editorial overview: Geologists do much more than identify rocks or interpret Earth's past. They also organize the underground assumptions needed for resource exploration, civil planning, and hazard assessment. Their work does not end with desk analysis; it depends on linking outcrops, terrain, groundwater, and prior reports into practical judgment.

### Governor

- URL: https://ai-job-risk.net/jobs/governor
- Slug: governor
- Industry: government.
- Current score: score 10, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace governors? Data organization and scenario comparison automate, but deciding who bears a policy's cost stays a human responsibility.
- Editorial overview: Governors do much more than approve government paperwork. They decide policy priorities for an entire prefecture, oversee crisis response, allocate budgets, coordinate with the national government, and connect with municipalities. They carry the political responsibility of deciding what to fund first and what to postpone while balancing the interests of a large region.

### Graphic Designer

- URL: https://ai-job-risk.net/jobs/graphic-designer
- Slug: graphic-designer
- Industry: creative.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace graphic designers? Image generation and draft layouts are fast now, but information hierarchy and brand tone still need a person.
- Editorial overview: A graphic designer is more than someone who decorates things. The role is about deciding what should be shown and in what order, based on information hierarchy, eye flow, brand tone, and the characteristics of the medium. The responsibility is not only for visual appeal, but also for communication itself.

### Historian

- URL: https://ai-job-risk.net/jobs/historian
- Slug: historian
- Industry: education.
- Current score: score 28, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace historians? Timelines and literature summaries automate first, but source criticism and reading what an omission means do not.
- Editorial overview: Historians do not simply summarize past events. They read sources, interpret events within the values and context of the time, and explain why those events happened. The profession is valued less for listing facts and more for the quality of source criticism and contextual interpretation.

### Hotel Manager

- URL: https://ai-job-risk.net/jobs/hotel-manager
- Slug: hotel-manager
- Industry: hospitality.
- Current score: score 32, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace hotel managers? Pricing, demand forecasting and review analysis automate; complaints and floor coordination stay on the ground.
- Editorial overview: A hotel manager does more than fill rooms. The role is to keep the whole operation working by watching guest experience, staff allocation, issue response, and revenue management at the same time. It is a role that sits directly at the intersection of the floor and management.

### Housekeeper

- URL: https://ai-job-risk.net/jobs/housekeeper
- Slug: housekeeper
- Industry: hospitality.
- Current score: score 24, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace housekeepers? Robotic cleaning and progress tracking automate, but reading a room for smell, damp and equipment trouble does not.
- Editorial overview: A housekeeper is more than someone who cleans a room. The role is to bring a room into a state where it works as part of a guest experience. It supports final quality not only through cleanliness, but through how amenities are arranged, how the room smells, and whether anything feels off.

### HR Specialist

- URL: https://ai-job-risk.net/jobs/hr-specialist
- Slug: hr-specialist
- Industry: consulting.
- Current score: score 53, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace HR specialists? Policy search and FAQ drafts automate, but applying one rule to illness, childcare, or conflict needs a person.
- Editorial overview: An HR specialist does more than process procedures and run systems. The role sits between rules and individual circumstances to help employees work with stability, across hiring, transfers, evaluations, attendance, labor matters, and policy communication. It is both an administrative role and one with responsibility for operational judgment.

### Human Resources Manager

- URL: https://ai-job-risk.net/jobs/human-resources-manager
- Slug: human-resources-manager
- Industry: consulting.
- Current score: score 45, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace HR managers? Attrition dashboards and policy drafts automate; deciding where hiring money goes and when to intervene does not.
- Editorial overview: An HR manager does more than supervise HR operations. The role is about deciding which people issues should come first and where resources should be allocated across hiring, staffing, evaluations, development, labor risk, and organizational culture. Compared with case-by-case handling, the responsibility for organizational line-drawing and decision-making is much greater.

### HVAC Technician

- URL: https://ai-job-risk.net/jobs/hvac-technician
- Slug: hvac-technician
- Industry: construction.
- Current score: score 24, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace HVAC technicians? Error-code candidates and part lists are quick now, but isolating overlapping causes on site is still human.
- Editorial overview: HVAC technicians do far more than replace equipment. They balance temperature, humidity, airflow, ductwork, refrigerant, controls, and building usage conditions so that comfort, energy efficiency, and maintainability can coexist. On new projects as well as in fault diagnosis, maintenance, retrofits, and operational improvement, field judgment matters.

### Illustrator

- URL: https://ai-job-risk.net/jobs/illustrator
- Slug: illustrator
- Industry: creative.
- Current score: score 75, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace illustrators? Rough ideation and reference generation are automating, but matching style to purpose and reading the brief are not.
- Editorial overview: An illustrator is more than someone who makes pretty pictures. The role is about reading the intent behind text or a concept and deciding what kind of atmosphere, metaphor, or visual framing will make it land. Illustration acts not only as an image, but as a device that converts meaning into something visible.

### Industrial Engineer

- URL: https://ai-job-risk.net/jobs/industrial-engineer
- Slug: industrial-engineer
- Industry: manufacturing.
- Current score: score 54, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace production engineers? Takt-time analysis and simulation automate, but finding the real bottleneck happens on the factory floor.
- Editorial overview: Production engineers do a great deal more than build processes. Their job is to design the conditions that keep an entire line running stably by balancing equipment capability, human motion, quality requirements, safety, logistics constraints, and cost. They are responsible not only for proposing improvements, but for turning them into systems that actually take root on the shop floor.

### Instructional Designer

- URL: https://ai-job-risk.net/jobs/instructional-designer
- Slug: instructional-designer
- Industry: education.
- Current score: score 51, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace instructional designers? Draft content and quiz banks come fast, but sequencing and valid assessment design stay human work.
- Editorial overview: Instructional designers do much more than make materials look clear and polished. Working backward from a learning goal, they decide which activities to include, how to confirm mastery, and what kind of learning experience will lead to retention and behavior change. They play an important role not only in schools, but also in corporate training, e-learning, and onboarding design.

### Insurance Agent

- URL: https://ai-job-risk.net/jobs/insurance-agent
- Slug: insurance-agent
- Industry: finance.
- Current score: score 55, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace insurance agents? Policy comparison tables are close to automated already; finding what a household truly needs covered is not.
- Editorial overview: Insurance agents do far more than explain products. Their job is to listen to a person’s family situation, income, future anxieties, and attitude toward protection, then help them organize which risks they actually need to prepare for and how. It may look like selling a numerical product, but in practice it is often closer to designing peace of mind.

### Insurance Underwriter

- URL: https://ai-job-risk.net/jobs/insurance-underwriter
- Slug: insurance-underwriter
- Industry: finance.
- Current score: score 74, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace insurance underwriters? Initial screening and risk scoring automate, but borderline cases and endorsement terms still need judgment.
- Editorial overview: Insurance underwriters do much more than review application details. Their job is to decide whether a policy should be accepted at all, under what terms it can be accepted, and whether the premium level and endorsements are appropriate. They balance profitability and risk by looking at claim rates, applicant attributes, disclosures, medical history, contract terms, and the overall portfolio mix.

### Interior Designer

- URL: https://ai-job-risk.net/jobs/interior-designer
- Slug: interior-designer
- Industry: creative.
- Current score: score 41, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace interior designers? Renders and mood boards take minutes now, but circulation, fire rules, and construction fit need a designer.
- Editorial overview: An interior designer is more than someone who makes a room look stylish. The role is about designing a spatial experience by considering movement flow, dwell time, psychological comfort, building conditions, budget, and construction constraints. It carries responsibility for making appearance and usability work at the same time.

### Interpreter

- URL: https://ai-job-risk.net/jobs/interpreter
- Slug: interpreter
- Industry: media.
- Current score: score 60, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace interpreters? Machine interpreting already covers travel and formulaic talk; high-stakes rooms still need a person reading them.
- Editorial overview: Interpreters do a great deal more than convert spoken words into another language on the spot. Their role is to act as a bridge so that meaning is conveyed correctly while reading the purpose of the conversation, the relationship between the speakers, the level of tension, and what is being implied but not said directly. In settings such as business negotiations, meetings, medical care, and formal negotiations, preserving the intent of the situation matters more than word-for-word substitution.

### Investment Analyst

- URL: https://ai-job-risk.net/jobs/investment-analyst
- Slug: investment-analyst
- Industry: finance.
- Current score: score 53, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace investment analysts? Earnings summaries and valuation models automate; the thesis and the exit discipline remain yours.
- Editorial overview: Investment analysts do far more than summarize disclosures and calculate valuations. They build and challenge investment theses, evaluate management quality and business structure, compare market expectations with their own view, and decide what downside matters enough to change or exit a position.

### Investment Banker

- URL: https://ai-job-risk.net/jobs/investment-banker
- Slug: investment-banker
- Industry: finance.
- Current score: score 40, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace investment bankers? DCF work, comps and pitch book drafting automate fast, but structuring and closing a deal still needs a person.
- Editorial overview: Investment bankers help bring large transactions to completion, capital raises, M&A, business sales, and restructurings, by connecting companies, investors, management teams, and legal and accounting advisors. The job is often associated with numbers, but in practice its center of gravity is deal design and stakeholder coordination.

### IT Support Specialist

- URL: https://ai-job-risk.net/jobs/it-support-specialist
- Slug: it-support-specialist
- Industry: technology.
- Current score: score 70, weekly change +1, week 2026-09-02.
- Latest score explanation: Persistent agents and local chatbot deployment strengthen AI’s ability to resolve common IT tickets, guide troubleshooting, and execute repeatable support steps without constant human prompting. Because many IT support tasks are digital and process-driven, recent developments modestly raise replacement risk.
- Editorial description: Will AI replace IT support specialists? Chatbots answer known questions, but vague reports with overlapping causes still need someone to sort.
- Editorial overview: IT support specialists do a great deal more than answer questions. Their job is to identify what is actually stopping work, restore service as quickly as possible, and improve operations so the same issue is less likely to recur. Because they deal across accounts, devices, networks, SaaS, permissions, and security operations, the role requires both technical knowledge and business understanding.

### Journalist

- URL: https://ai-job-risk.net/jobs/journalist
- Slug: journalist
- Industry: media.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace journalists? Press-conference summaries and headline options automate; on-the-ground verification and reporting design do not.
- Editorial overview: Journalists are not simply people who gather events from the world and turn them into articles. The essence of the role is deciding what matters as news, going to primary sources, verifying facts, and delivering them to readers in a socially meaningful form. The value of the job is determined less by the amount of information collected and more by the quality of reporting and the quality of the questions being asked.

### Judge

- URL: https://ai-job-risk.net/jobs/judge
- Slug: judge
- Industry: legal.
- Current score: score 11, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace judges? More than 60 percent of federal judges now use an AI tool in chambers, a March 2026 survey found, but rulings stay human.
- Editorial overview: Judges do much more than apply legal provisions mechanically. They weigh the credibility of evidence, identify what matters most in competing claims, and express a decision in words while preserving procedural fairness. By 2026, tools built specifically for judicial workflows, such as CoCounsel Legal on the Westlaw platform, sit inside a large share of chambers, but a March 2026 survey of federal judges found only about 22 percent use one weekly or daily.

### Laboratory Technician

- URL: https://ai-job-risk.net/jobs/laboratory-technician
- Slug: laboratory-technician
- Industry: healthcare.
- Current score: score 52, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace laboratory technicians? Waveform triage and abnormal-value extraction automate, but judging sample validity and retesting does not.
- Editorial overview: Laboratory technicians do far more than produc numbers. Their work is to support test quality that can actually be trusted in clinical care through specimen handling, pre-analytic preparation, equipment management, validation of results, decisions about retesting, and the handling of abnormal values. Their responsibility is not only to generate results, but to protect the conditions under which those results are reliable.

### Lawyer

- URL: https://ai-job-risk.net/jobs/lawyer
- Slug: lawyer
- Industry: legal.
- Current score: score 40, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace lawyers? Red-flag contract review and precedent search automate, but negotiation and accountable advice stay with the lawyer.
- Editorial overview: Lawyers do far more than answer legal questions. They translate a client's real-world situation into legal issues, show the available options and risks, and adapt their role depending on whether the work involves preventive contracts, negotiations after a dispute, litigation, or internal advice.

### Legal Assistant

- URL: https://ai-job-risk.net/jobs/legal-assistant
- Slug: legal-assistant
- Industry: legal.
- Current score: score 68, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace legal assistants? Clause extraction and deadline spotting automate first; version control and approval flows stay human work.
- Editorial overview: Legal assistants do much more than prepar contracts and filing documents. They keep legal operations running without gaps by supporting documents, deadlines, approval flows, version control, formatting adjustments, and follow-up requests across stakeholders.

### Librarian

- URL: https://ai-job-risk.net/jobs/librarian
- Slug: librarian
- Industry: education.
- Current score: score 40, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace librarians? Title search and recommendations automate easily, but shaping a question users cannot yet articulate does not.
- Editorial overview: Librarians do far more than plac books on shelves. They are professionals who deliver the right materials to the right people in the right way. Their role includes classification, reference support, collection development, user guidance, and building the learning infrastructure of a community.

### Loan Officer

- URL: https://ai-job-risk.net/jobs/loan-officer
- Slug: loan-officer
- Industry: finance.
- Current score: score 60, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace loan officers? Scoring on income, debt ratio and collateral automates; job changes, self-employment and missing documents do not.
- Editorial overview: Loan officers do much more than take in application details and pass them on for review. Their job is to improve the chances that a loan can actually be approved by aligning the borrower’s situation, the repayment plan, the required documents, and the lending institution’s standards. In areas like mortgages and business lending, that means understanding both the borrower’s anxieties and the lender’s logic.

### Logistics Coordinator

- URL: https://ai-job-risk.net/jobs/logistics-coordinator
- Slug: logistics-coordinator
- Industry: logistics.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace logistics coordinators? Tracking, ETA forecasts, and delay alerts automate; deciding which client to protect in a delay does not.
- Editorial overview: Logistics coordinators do far more than arrange shipments. They keep the entire flow from stalling by coordinating inbound and outbound schedules, warehouse status, vehicle allocation, deadlines, customs and document requirements, and communication with business partners. Their responsibility involves more than creating a plan; it also involves rebuilding it when it breaks.

### Machine Learning Engineer

- URL: https://ai-job-risk.net/jobs/machine-learning-engineer
- Slug: machine-learning-engineer
- Industry: technology.
- Current score: score 16, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace machine learning engineers? Baseline models get easy; data quality, evaluation design and drift monitoring stay human work.
- Editorial overview: Machine learning engineers do far more than train models. Their role is to build a system that extracts value from data and keeps running stably in production. That means dealing with features, training data, evaluation metrics, serving, monitoring, and drift management as one connected workflow. Research alone is not enough, and implementation alone is not enough either.

### Management Consultant

- URL: https://ai-job-risk.net/jobs/management-consultant
- Slug: management-consultant
- Industry: consulting.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace management consultants? Market research and meeting materials draft themselves now; framing the real executive issue does not.
- Editorial overview: Management consultants do far more than produc slides. They organize ambiguous management issues, build hypotheses, involve stakeholders, and shape choices into a form that leadership can actually decide on. Their responsibility is both analysis and creating conditions in which executives can move.

### Manufacturing Engineer

- URL: https://ai-job-risk.net/jobs/manufacturing-engineer
- Slug: manufacturing-engineer
- Industry: manufacturing.
- Current score: score 42, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace manufacturing engineers? Yield analysis and parameter suggestions automate, but keeping mass production stable does not. See the split.
- Editorial overview: Manufacturing engineers do far more than keep equipment running. Their job is to design the conditions that make a process hold together by balancing product specifications, processing conditions, equipment capability, yield, quality defects, and production stability. They are responsible for closing the gap between what works in prototyping and what can survive in mass production.

### Market Research Analyst

- URL: https://ai-job-risk.net/jobs/market-research-analyst
- Slug: market-research-analyst
- Industry: marketing.
- Current score: score 62, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace market research analysts? Desk research and survey summaries automate, but designing what to verify decides whether it is useful.
- Editorial overview: Market research analysts do far more than tabulate survey results. Their role is to clarify the decision that needs to be made, design the kind of research that can move the organization closer to an answer, interpret both quantitative and qualitative findings, and turn them into business insight. The quality of the work depends on everything from method selection and question design to sample interpretation and how results are read.

### Marketing Manager

- URL: https://ai-job-risk.net/jobs/marketing-manager
- Slug: marketing-manager
- Industry: marketing.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace marketing managers? Dashboards and budget drafts automate, but the trade-off between acquisition and brand stays a human call.
- Editorial overview: A marketing manager is not simply the supervisor of campaign specialists. The role is to decide which markets to target, which channels deserve investment, and what team structure will be used to produce business results. They carry responsibility for the overall priorities of marketing, including budget, headcount, outside partners, brand direction, and coordination with sales.

### Marketing Specialist

- URL: https://ai-job-risk.net/jobs/marketing-specialist
- Slug: marketing-specialist
- Industry: marketing.
- Current score: score 68, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace marketing specialists? Copy ideas and report summaries mass-produce cheaply, but deciding what to stop funding does not.
- Editorial overview: Marketing specialists are not simply people who run ads or update social media. Their role is to clarify whose problem a product or service solves and what value it offers, translate that into initiatives, and then improve based on results. Because the work links market understanding, messaging design, channel operation, and internal coordination, the real difference comes less from the amount of work done and more from the quality of judgment.

### Mayor

- URL: https://ai-job-risk.net/jobs/mayor
- Slug: mayor
- Industry: government.
- Current score: score 11, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace mayors? Procedures and document preparation automate, but reading local pain points and holding public trust stays human work.
- Editorial overview: Mayors do a great deal more than speake as the face of a municipality. They oversee public services, the local economy, welfare, education, infrastructure, and disaster response at close range to the community. They must understand both the problems residents feel directly and the realities inside city hall, then decide what to advance first under tight budgets.

### Mechanic

- URL: https://ai-job-risk.net/jobs/mechanic
- Slug: mechanic
- Industry: manufacturing.
- Current score: score 30, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace industrial mechanics? Vibration monitoring and inspection logs automate, but isolating a fault and stopping the line does not.
- Editorial overview: Industrial mechanics do much more than replace broken parts. Their job is to listen to equipment sounds, watch for vibration, wear, heat, and behavioral quirks, then decide what the real cause is and how far the machinery should be shut down for repair. It is a role that carries responsibility both for keeping equipment running and for deciding when it must be stopped.

### Mechanical Engineer

- URL: https://ai-job-risk.net/jobs/mechanical-engineer
- Slug: mechanical-engineer
- Industry: manufacturing.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace mechanical engineers? Shape options and past-design search speed up; tolerance stack-up and failure modes stay human calls.
- Editorial overview: Mechanical engineers do much more than draw plans. They design mechanisms that can actually be manufactured and used over time while balancing strength, weight, manufacturability, cost, assembly, maintainability, and safety. Their responsibility involves more than making something look feasible on paper; it also involves thinking through how it may fail in the real world.

### Medical Assistant

- URL: https://ai-job-risk.net/jobs/medical-assistant
- Slug: medical-assistant
- Industry: healthcare.
- Current score: score 44, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace medical assistants? Intake formatting and appointment confirmation speed up; guiding anxious patients through a clinic does not.
- Editorial overview: Medical assistants do a great deal more than help at reception. Their work is to keep clinical settings running smoothly by supporting pre-visit preparation, patient guidance, intake assistance, record support, supply management, and the workflow that lets doctors and nurses move efficiently. Even without making diagnoses, they have a direct impact on how well the care environment functions.

### Meteorologist

- URL: https://ai-job-risk.net/jobs/meteorologist
- Slug: meteorologist
- Industry: science.
- Current score: score 47, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace meteorologists? Baseline forecast quality keeps improving, but terrain correction and warning communication still need people.
- Editorial overview: Meteorologists do a great deal more than state whether it will rain or shine. Their role is to explain weather by interpreting observations and numerical models and to judge how specific developments will affect transportation, agriculture, energy, and disaster prevention at the regional level.

### Military Officer

- URL: https://ai-job-risk.net/jobs/military-officer
- Slug: military-officer
- Industry: public-service.
- Current score: score 10, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace military officers? Surveillance and route proposals automate, but discipline, morale, and final command judgment stay human.
- Editorial overview: Military officers do far more than carry out orders. They combine situational awareness, unit operations, logistics, safety, discipline, and field judgment to move people and equipment toward mission success. From peacetime training to wartime response, they are responsible for maintaining control in ambiguous conditions.

### Miner

- URL: https://ai-job-risk.net/jobs/miner
- Slug: miner
- Industry: manufacturing.
- Current score: score 48, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace miners? Haul routing and safety records automate first, but reading the ground and calling a stop stays a human job on site.
- Editorial overview: Miners do much more than dig. They keep extraction moving while watching ground conditions, heavy equipment movement, ventilation, safe distances, collapse risk, and work order so danger can be avoided. This is a job where safety decisions tied directly to human life matter as much as productivity.

### Mobile App Developer

- URL: https://ai-job-risk.net/jobs/mobile-app-developer
- Slug: mobile-app-developer
- Industry: technology.
- Current score: score 78, weekly change +1, week 2026-09-02.
- Latest score explanation: OpenAI’s persistent coding agent work suggests AI can stay on task longer across debugging, refactoring, and feature implementation, raising pressure on app development workflows. The increase is small because integration, product tradeoffs, and quality control still require substantial human engineering input.
- Editorial description: Will AI replace mobile app developers? Screen scaffolding and API glue code generate fast, but OS updates, permissions and store review do not.
- Editorial overview: Mobile app developers create experiences people use on smartphones while working within device limitations and OS-specific behavior. The job goes far beyond building screens. It also involves communication conditions, notifications, background behavior, device-performance differences, app-store review, and crash handling. Compared with the web, the usage environment is more constrained, so it is especially important to think about how the product will be used in reality.

### Model

- URL: https://ai-job-risk.net/jobs/model
- Slug: model
- Industry: entertainment.
- Current score: score 55, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace models? Standard ecommerce assets and mass variations go first; brand-face, live and physical-expression work is harder to replace.
- Editorial overview: A model does more than wear clothes or hold products in front of a camera. The role is to communicate how a brand wants to be seen through physical expression, working in sync with the photo or video team to make the impression of a product or concept concrete.

### Museum Curator

- URL: https://ai-job-risk.net/jobs/museum-curator
- Slug: museum-curator
- Industry: education.
- Current score: score 24, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace museum curators? AI drafts captions and organizes collection data, but deciding what to exhibit and what to leave out does not.
- Editorial overview: Museum curators do much more than stor collections. They research the meaning of objects and materials, then communicate that meaning to society through exhibitions and interpretation. Their role spans preservation, research, exhibition design, and public education, and they are responsible for deciding how cultural assets should be presented.

### Music Producer

- URL: https://ai-job-risk.net/jobs/music-producer
- Slug: music-producer
- Industry: entertainment.
- Current score: score 57, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace music producers? Demos and reference work speed up, but choosing direction and protecting an artist's identity stays human.
- Editorial overview: A music producer does more than collect songs and demos. The work involves defining an artist’s direction, deciding roles across the production team, judging what should be developed or discarded, and connecting the music itself to how it will be presented and sold.

### Network Engineer

- URL: https://ai-job-risk.net/jobs/network-engineer
- Slug: network-engineer
- Industry: technology.
- Current score: score 49, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace network engineers? Standard configs and log summaries template well; isolating failures in a real dependency mess does not.
- Editorial overview: Network engineers do much more than enter device settings. Their role is to design structures where communication keeps flowing, remains secure, and stays easy to troubleshoot during outages. They have to balance speed, redundancy, security, and operability across routers, switches, VPNs, firewalls, and cloud connectivity.

### Nurse

- URL: https://ai-job-risk.net/jobs/nurse
- Slug: nurse
- Industry: healthcare.
- Current score: score 15, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace nurses? Nursing notes, vitals summaries and handoffs get drafted faster, but noticing small changes at the bedside does not.
- Editorial overview: Nurses do far more than carry out instructions. Their job is to observe subtle changes in patient condition at close range and support a state in which treatment can continue through daily care, procedures, family communication, physician reporting, and coordination with other professionals. They are important not only in supporting medical acts but also in noticing anxiety and the realities of a patient's life.

### Office Clerk

- URL: https://ai-job-risk.net/jobs/office-clerk
- Slug: office-clerk
- Industry: operations.
- Current score: score 87, weekly change +1, week 2026-09-02.
- Latest score explanation: Office-clerk duties such as data handling, routing requests, updating records, and basic coordination are well aligned with the persistent-agent trend in this week’s news. AI’s improving ability to execute multi-step digital workflows makes this role slightly more exposed than last week.
- Editorial description: Will AI replace office clerks? BLS projects a 7% decline in general office clerk jobs by 2034. See which tasks automate first and which stay human.
- Editorial overview: Office clerks do much more than update forms and handle general paperwork. They support daily business operations by keeping records current, organizing documents, responding to routine inquiries, checking figures and fields, and preventing small administrative errors from spreading.

### Operations Manager

- URL: https://ai-job-risk.net/jobs/operations-manager
- Slug: operations-manager
- Industry: consulting.
- Current score: score 47, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace operations managers? KPI dashboards and staffing drafts automate; deciding what to stop when problems pile up still does not.
- Editorial overview: An operations manager does more than supervise the frontline. The role is about deciding which problems to prioritize and which operational changes to allow each day while balancing staffing, quality, deadlines, cost, safety, and customer impact. The responsibility lies less in watching numbers than in drawing the line that keeps operations from stopping.

### Paralegal

- URL: https://ai-job-risk.net/jobs/paralegal
- Slug: paralegal
- Industry: legal.
- Current score: score 76, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace paralegals? First-pass document review automates well, but spotting where a summary is unreliable and evidence missing does not.
- Editorial overview: Paralegals support lawyers and legal teams by gathering materials aligned to specific issues and organizing evidence and legal information into forms that help judgment. Their value lies not in collecting documents alone, but in seeing what relates to the dispute and where gaps still remain.

### Paramedic

- URL: https://ai-job-risk.net/jobs/paramedic
- Slug: paramedic
- Industry: healthcare.
- Current score: score 14, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace paramedics? Triage support and record drafting automate, but scene safety and severity calls in the first minutes do not.
- Editorial overview: Paramedics do much more than transport patients. Their work is to assess danger upon arrival, observe the patient's condition, deliver treatment under severe time constraints, decide on the receiving destination, and communicate with family and bystanders. They carry responsibility for life-affecting initial response in unstable environments outside the hospital.

### Pharmacist

- URL: https://ai-job-risk.net/jobs/pharmacist
- Slug: pharmacist
- Industry: healthcare.
- Current score: score 50, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace pharmacists? Interaction checks and counseling drafts automate, but judging whether a prescription holds up in daily life does not.
- Editorial overview: Pharmacists do much more than hand out medicine. Their work is to judge whether a prescribed drug regimen is truly safe and sustainable by considering the prescription itself, interactions, renal and liver function, adherence status, and the patient's lifestyle. Through prescription review, physician queries, medication counseling, and home-care support, they protect the safety of treatment.

### Photographer

- URL: https://ai-job-risk.net/jobs/photographer
- Slug: photographer
- Industry: creative.
- Current score: score 28, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace photographers? Retouching, background removal and rough selection automate; timing and trust with a subject stay human work.
- Editorial overview: A photographer does more than press the shutter. The role is about deciding what should be shown, then shaping light, distance, expression, timing, and context into a single frame. The responsibility lies less in operating equipment than in choosing the moment that matters.

### Physicist

- URL: https://ai-job-risk.net/jobs/physicist
- Slug: physicist
- Industry: science.
- Current score: score 25, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace physicists? Calculation and literature review accelerate, but choosing what a model can safely neglect does not.
- Editorial overview: Physicists do far more than manipulate equations. Their role is to decide what can be simplified, what counts as a valid approximation, and how theory, experiment, measurement, and simulation should be linked in order to explain a phenomenon.

### Pilot

- URL: https://ai-job-risk.net/jobs/pilot
- Slug: pilot
- Industry: transportation.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace pilots? Flight planning, fuel forecasting and anomaly alerts automate, but abnormal-situation calls and crew alignment do not.
- Editorial overview: Pilots are not simply the people who move the controls. They are responsible for making safe operations possible by weighing aircraft condition, weather, airport conditions, fuel, and crew status before, during, and after flight.

### Plumber

- URL: https://ai-job-risk.net/jobs/plumber
- Slug: plumber
- Industry: construction.
- Current score: score 11, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace plumbers? BLS projects about 6 percent growth to 2033 and 43,300 openings a year. AI drafts the quote; the wall needs a plumber.
- Editorial overview: Plumbers do far more than connect pipes. They make installation decisions based on water supply, drainage, sanitation, gas, slope, serviceability, and leak risk, deciding where piping should run and how future maintenance can be made easier. By 2026, much of the office work around that decision, quoting, scheduling, and documenting a job, runs through software like ServiceTitan, Jobber, or Housecall Pro. The physical judgment on site has not moved to a screen.

### Police Officer

- URL: https://ai-job-risk.net/jobs/police-officer
- Slug: police-officer
- Industry: public-service.
- Current score: score 14, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace police officers? Video analysis and record work shrink, but defusing a scene without escalating it stays human. See the split.
- Editorial overview: Police officers do much more than enforc rules. They assess situations in the field, prevent crime and accidents before they escalate, and maintain public safety while dealing directly with victims and local residents. Patrols, interviews, initial investigations, traffic response, and community coordination are all part of the job.

### Politician

- URL: https://ai-job-risk.net/jobs/politician
- Slug: politician
- Industry: government.
- Current score: score 12, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace politicians? Speeches, posts and issue mapping automate easily, but choosing which conflict to own cannot be delegated.
- Editorial overview: Politicians do far more than read policy proposals. Their work is to give language to social conflict, frustration, and hope, gather support, and move institutions. They operate in the broad field of politics itself, including elections, supporter relations, position-building in assemblies, and bargaining with other parties and factions.

### Power Plant Operator

- URL: https://ai-job-risk.net/jobs/power-plant-operator
- Slug: power-plant-operator
- Industry: energy.
- Current score: score 51, weekly change -1, week 2026-09-02.
- Latest score explanation: This week’s warnings about attacks on water systems and broader AI security failures slightly reduce near-term replacement risk for critical infrastructure operators. Real-time plant oversight demands trusted human judgment and accountable intervention when cyber-physical systems behave unexpectedly.
- Editorial description: Will AI replace power plant operators? Monitoring and first-line detection automate, but judging how dangerous a sign is stays human responsibility.
- Editorial overview: Power plant operators do far more than watch control panels. They monitor plant condition, fuel conditions, grid demands, safety procedures, and warning signs, then decide whether to continue, stop, or change priorities. They are on the front line of balancing stable supply with safety.

### Procurement Specialist

- URL: https://ai-job-risk.net/jobs/procurement-specialist
- Slug: procurement-specialist
- Industry: consulting.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace procurement specialists? Quote comparison and contract drafts automate, but supplier negotiation and supply risk stay human.
- Editorial overview: A procurement specialist does much more than collect quotes. The role is about deciding what to buy, and from whom, while balancing price, lead time, quality, supply stability, contract terms, and supplier risk. The job is responsible not only for unit cost, but also for keeping supply from breaking.

### Product Manager

- URL: https://ai-job-risk.net/jobs/product-manager
- Slug: product-manager
- Industry: technology.
- Current score: score 33, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace product managers? PRDs and competitor lists draft themselves, but explaining what gets cut and why is still a person's job.
- Editorial overview: A product manager is not simply the person who writes specs. The core of the role is deciding where limited development resources should be used. Product managers compare customer requests, business goals, technical constraints, operational burden, and revenue impact, then decide what should be done now and what should be left out.

### Professor

- URL: https://ai-job-risk.net/jobs/professor
- Slug: professor
- Industry: education.
- Current score: score 18, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace professors? Literature summaries and lecture drafts speed up, but setting research themes and supervising students do not.
- Editorial overview: Professors do far more than give lectures. They define research themes, update the stock of knowledge, supervise students, and mobilize resources inside and outside the university while carrying responsibility for both education and research. Their value comes from the combined work of teaching, supervision, writing, conferences, funding, collaborative research, and institutional management.

### Programmer

- URL: https://ai-job-risk.net/jobs/programmer
- Slug: programmer
- Industry: technology.
- Current score: score 77, weekly change +1, week 2026-09-02.
- Latest score explanation: OpenAI’s reported persistent agent and the broader focus on autonomous coding workflows are direct evidence of stronger automation for writing, testing, and iterating software. The score rises only one point because this same week’s governance and reliability issues show why human supervision still matters in production development.
- Editorial description: Will AI replace programmers? Well-specified implementation is already going to code assistants; requirements, design and operations are not.
- Editorial overview: Programmers do far more than simply convert specifications into code. In practice, they create value across an end-to-end flow that includes interpreting requirements, designing systems, implementing solutions, reviewing code, testing, and operating software. In particular, professionals who can decide what should be absorbed by systems and where human judgment should remain have a different kind of market value from simple coding workers.

### Project Manager

- URL: https://ai-job-risk.net/jobs/project-manager
- Slug: project-manager
- Industry: consulting.
- Current score: score 43, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace project managers? Status reports and task lists automate, but scope creep, dependencies, and stalled decisions need a person.
- Editorial overview: A project manager does more than manage a schedule. The role is about moving a project forward with limited time and resources while balancing objectives, scope, priorities, risks, and stakeholders. More than noticing delay, the profession carries responsibility for deciding when direction has to change.

### Proofreader

- URL: https://ai-job-risk.net/jobs/proofreader
- Slug: proofreader
- Industry: media.
- Current score: score 76, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace proofreaders? Typos and inconsistent sentence endings are already machine work; judging context and publication quality is not.
- Editorial overview: Proofreaders do far more than find typos. They are responsible for the precision of writing through work such as unifying style rules, spotting distortions in meaning, catching mismatches between subjects and predicates, checking consistency in quotations and notes, and ensuring compliance with publication standards. It is a detail-oriented role, but in practice it also serves as the final gate that protects publication quality.

### Property Manager

- URL: https://ai-job-risk.net/jobs/property-manager
- Slug: property-manager
- Industry: real-estate.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace property managers? Monthly reports and inquiry handling standardize well; deciding which repair protects value first does not.
- Editorial overview: Property managers do a great deal more than leas out buildings. They are responsible for balancing occupancy, rent levels, repairs, tenant satisfaction, and owner returns. In practice, the job is both about putting out day-to-day fires and about making repeated decisions that keep a building valuable over time.

### Prosecutor

- URL: https://ai-job-risk.net/jobs/prosecutor
- Slug: prosecutor
- Industry: legal.
- Current score: score 24, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace prosecutors? Statement comparison and case-law search automate, but charging decisions and theories of proof stay with people.
- Editorial overview: Prosecutors do a great deal more than lin up collected evidence. They decide how far facts can actually be proven, whether charges should be brought, and in what order arguments and evidence should be presented in court. It is a role that brings together evidence assessment, procedure, victim impact, and social consequences into one weighty judgment.

### Psychiatrist

- URL: https://ai-job-risk.net/jobs/psychiatrist
- Slug: psychiatrist
- Industry: healthcare.
- Current score: score 16, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace psychiatrists? Interview summaries and drug searches automate, but risk assessment and the treatment relationship stay human.
- Editorial overview: Psychiatrists do much more than assign diagnoses and prescribe medication. Their work is to integrate what is said in the interview, emotional fluctuation, self-harm or harm-to-others risk, daily functioning, family relationships, and the influence of physical illness to decide what should be supported first. They work across diagnosis, medication, environmental adjustment, and multidisciplinary coordination at the same time.

### Psychologist

- URL: https://ai-job-risk.net/jobs/psychologist
- Slug: psychologist
- Industry: healthcare.
- Current score: score 12, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace psychologists? Session summaries and worksheets automate first, but assessment inside a real conversation stays human work.
- Editorial overview: Psychologists do far more than listen to people's problems. Their work is to use interviews, psychological testing, behavioral observation, and life-context analysis to understand what a person is struggling with and to build a support direction from that understanding. The role involves more than receiving what is said; it also involves judging what lies at the center of the problem and what support resources are needed.

### QA Engineer

- URL: https://ai-job-risk.net/jobs/qa-engineer
- Slug: qa-engineer
- Industry: technology.
- Current score: score 85, weekly change +1, week 2026-09-02.
- Latest score explanation: Persistent AI agents improve coverage of repetitive QA tasks such as regression checks, bug reproduction, and test generation across longer workflows. Given this week’s coding-agent momentum, QA work remains among the most exposed digital occupations and edges upward.
- Editorial description: Will AI replace QA engineers? Gartner expects a third of enterprise applications to run on agentic AI by 2028, yet release calls stay human.
- Editorial overview: QA engineers do a great deal more than run tests. Their job is to define what quality means, decide where defects are most likely to hide, and determine at what point a release should be stopped. By 2026, agentic coding and testing tools sit inside a large share of that daily work, drafting cases, healing broken scripts, and summarizing results, while the decision about how quality should be protected still sits with a person.

### Quality Assurance Specialist

- URL: https://ai-job-risk.net/jobs/quality-assurance-specialist
- Slug: quality-assurance-specialist
- Industry: manufacturing.
- Current score: score 60, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace quality assurance specialists? Defect trends and reports automate, but the call to ship or hold a product stays with a person.
- Editorial overview: Quality assurance specialists do far more than find defects during inspection. They are responsible for drawing the line on quality by interpreting specifications, deciding whether products can ship, preventing recurrence, and reconciling customer requirements with shop-floor reality. Their job involves more than counting problems; it also involves deciding where production should stop.

### Radiologist

- URL: https://ai-job-risk.net/jobs/radiologist
- Slug: radiologist
- Industry: healthcare.
- Current score: score 59, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace radiologists? Detection marks, prior comparison, and templated reports automate; reading findings in clinical context does not.
- Editorial overview: Radiologists do far more than look at images. Their job is to interpret findings from CT, MRI, X-ray, ultrasound, and other modalities in light of the patient's symptoms, test results, medical history, and the purpose of the exam, and then turn those findings into information that other clinicians can actually use. They are responsible both for listing findings and for clarifying what matters and what is urgent.

### Real Estate Agent

- URL: https://ai-job-risk.net/jobs/real-estate-agent
- Slug: real-estate-agent
- Industry: real-estate.
- Current score: score 31, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace real estate agents? Comparison tables by price and layout generate instantly; surfacing a client's unspoken worry does not.
- Editorial overview: Real estate agents do far more than show property listings. They listen to a client’s budget, family situation, commute, future plans, and anxieties, then support the client’s decision-making through to contract. In both sales and rentals, the job often involves conversations close to life choices rather than just product explanation.

### Real Estate Broker

- URL: https://ai-job-risk.net/jobs/real-estate-broker
- Slug: real-estate-broker
- Industry: real-estate.
- Current score: score 41, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace real estate brokers? Listing materials and market comparisons automate, but holding a deal together through negotiation does not.
- Editorial overview: Real estate brokers play a broader role than ordinary property sales. They connect sellers and buyers, landlords and tenants, investors, lenders, and management companies while helping the transaction as a whole come together. Their work often includes condition adjustment, negotiation, information disclosure, and organizing conflicting interests across the deal.

### Receptionist

- URL: https://ai-job-risk.net/jobs/receptionist
- Slug: receptionist
- Industry: hospitality.
- Current score: score 80, weekly change +1, week 2026-09-02.
- Latest score explanation: Voice-first AI hardware and always-on assistant capabilities add to existing automation pressure on call handling, appointment intake, and front-desk information tasks. This week’s developments make receptionist workflows slightly more substitutable by conversational AI systems.
- Editorial description: Will AI replace receptionists? Reservation checks and identity verification automate, but reading a nervous visitor and routing them does not.
- Editorial overview: A receptionist does more than provide directions. The role is to understand a visitor's purpose and uncertainty in a very short time and guide them into the right path for the setting. As the first point of contact, reception shapes the impression of the whole place.

### Recruiter

- URL: https://ai-job-risk.net/jobs/recruiter
- Slug: recruiter
- Industry: consulting.
- Current score: score 65, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace recruiters? Sourcing, resume summaries, and scheduling automate, but defining the talent profile and aligning interviewers does not.
- Editorial overview: A recruiter does more than gather candidates. The role is about building a hiring process that brings in the right people for the organization by shaping job requirements, building the candidate pipeline, designing interviews, managing candidate experience, and aligning selection standards. The responsibility lies less in scheduling and more in assessment and alignment.

### Renewable Energy Technician

- URL: https://ai-job-risk.net/jobs/renewable-energy-technician
- Slug: renewable-energy-technician
- Industry: energy.
- Current score: score 31, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace renewable energy technicians? Monitoring flags anomalies, but on-site root-cause diagnosis and maintenance calls stay human.
- Editorial overview: Renewable energy technicians keep solar, wind, storage, and related equipment running safely, prevent failures, and stabilize generation output in the field. Their role does not end at installation. They maintain long-term usability through inspections, maintenance, grid connection work, and responses to environmental conditions.

### Research Assistant

- URL: https://ai-job-risk.net/jobs/research-assistant
- Slug: research-assistant
- Industry: science.
- Current score: score 55, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace research assistants? Literature search and first aggregation speed up, but sample condition and record quality still need care.
- Editorial overview: Research assistants are more than support staff who follow instructions. They help sustain reproducibility through sample management, records, literature organization, experiment preparation, and schedule coordination. It is an often invisible role, but one where poor records or sloppy condition control can damage an entire project.

### Retail Cashier

- URL: https://ai-job-risk.net/jobs/retail-cashier
- Slug: retail-cashier
- Industry: retail.
- Current score: score 80, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace cashiers? BLS expects cashier employment to fall roughly 10 percent from 2024 to 2034. See what checkout work still needs a person.
- Editorial overview: A cashier is the store's last human checkpoint before a customer leaves with a cart, and in 2026 that checkpoint looks different from a year ago. Some chains have added camera-based exit systems that confirm what's in a cart against a receipt in seconds; others have quietly pulled self-checkout lanes back out and restaffed registers with people instead. The job still runs on constant small judgment calls: what to do when a price won't scan, when a discount code looks wrong, or when a line backs up and someone in it is getting anxious.

### Retail Salesperson

- URL: https://ai-job-risk.net/jobs/retail-salesperson
- Slug: retail-salesperson
- Industry: retail.
- Current score: score 46, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace retail salespeople? Automated recommendation hits simple everyday goods hardest; fit, gifting, and hesitation still need a person.
- Editorial overview: Retail salespeople support customers in choosing products that fit their preferences and intended use. Their job involves more than explaining features; it also involves guiding selection in a way that also accounts for how the item will be used, what alternatives the customer is comparing, and whether the customer is likely to run into problems after purchase.

### Robotics Engineer

- URL: https://ai-job-risk.net/jobs/robotics-engineer
- Slug: robotics-engineer
- Industry: technology.
- Current score: score 25, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace robotics engineers? Control drafts and log analysis speed up, but sensor error, wear and field safety stay human judgment.
- Editorial overview: Robotics engineers do much more than write algorithms. Their role is to connect sensors, control, mechanics, electrical systems, and software into a system that can move safely in the real world. The gap between something that works in simulation and something that runs reliably in the field is large, so theoretical correctness alone is not enough.

### Sales Representative

- URL: https://ai-job-risk.net/jobs/sales-representative
- Slug: sales-representative
- Industry: marketing.
- Current score: score 36, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace sales representatives? Proposal emails and call summaries draft themselves; reading what a customer will not say does not.
- Editorial overview: Sales representatives do much more than explain products. Their role is to understand the prospect’s situation, organize the customer’s challenges, assemble the information needed for a decision, and move the process toward a contract or implementation. Product knowledge, customer understanding, proposal design, internal coordination, and closing are all connected, and what separates people is not how much they talk, but how well they can read a buyer’s anxiety and decision process.

### Scheduler

- URL: https://ai-job-risk.net/jobs/scheduler
- Slug: scheduler
- Industry: operations.
- Current score: score 95, weekly change +1, week 2026-09-02.
- Latest score explanation: Scheduling is one of the clearest fits for persistent agents because it relies on calendar access, reminders, follow-ups, and rules-based coordination across digital systems. This week’s agent developments directly strengthen end-to-end automation for the role’s core tasks, so it moves up slightly from an already high baseline.
- Editorial description: Will AI replace schedulers? Finding open slots and sending reminders automate, but priority calls and last-minute rescheduling still need people.
- Editorial overview: Schedulers do far more than drop meetings onto a calendar. They continually reorganize plans so that work does not jam up, balancing multiple stakeholders, deadlines, travel time, meeting preparation, priority cases, and the risk of delay. In some settings the work resembles executive support, while in others it is closer to operational command and control.

### School Counselor

- URL: https://ai-job-risk.net/jobs/school-counselor
- Slug: school-counselor
- Industry: education.
- Current score: score 16, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace school counselors? Record summaries and guardian messages automate, but catching bullying or school refusal early needs a person.
- Editorial overview: School counselors do a great deal more than listen to students’ concerns. They help organize the background behind a child’s stress and anxiety, connect the student, guardians, teachers, and school systems, and create conditions in which the student can learn safely. Their role includes preventive support, crisis response, and schoolwide coordination, not just one-on-one sessions.

### Security Guard

- URL: https://ai-job-risk.net/jobs/security-guard
- Slug: security-guard
- Industry: legal.
- Current score: score 31, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace security guards? Camera monitoring and entry logs automate first; deciding whether an alert is real harm stays a human call.
- Editorial overview: Security guards do far more than spot abnormalities. They decide what is truly dangerous, what can be handled as a routine issue, and when other people need to be moved or alerted. Their work supports daily safety through knowledge of facility flow, visitor movement, and emergency first response.

### SEO Specialist

- URL: https://ai-job-risk.net/jobs/seo-specialist
- Slug: seo-specialist
- Industry: marketing.
- Current score: score 76, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace SEO specialists? Keyword lists, outlines and audits generate fast, but choosing which queries and pages to bet on does not.
- Editorial overview: SEO specialists do far more than line up keywords and publish more articles. Their job is to understand search intent, design site architecture, content strategy, technical requirements, internal linking, and optimization priorities, then turn organic search traffic into business results. In practice, the role is highly cross-functional and often requires close coordination with editorial, engineering, design, and sales teams.

### Ship Captain

- URL: https://ai-job-risk.net/jobs/ship-captain
- Slug: ship-captain
- Industry: transportation.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace ship captains? Route candidates and fuel efficiency are calculated for you, but safety correction at sea and crew leadership are not.
- Editorial overview: Ship captains do a great deal more than keep a vessel moving. They oversee the safety of the voyage as a whole by weighing weather, sea conditions, cargo, crew, and port constraints. Safe marine operations depend on a long sequence of decisions made over extended time.

### Ship Engineer

- URL: https://ai-job-risk.net/jobs/ship-engineer
- Slug: ship-engineer
- Industry: transportation.
- Current score: score 28, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace ship engineers? Monitoring data and maintenance history automate, but deciding when to stop machinery at sea stays human work.
- Editorial overview: Ship engineers do far more than keep marine engines and auxiliary systems running. During long voyages, they must detect signs of abnormality early and decide when operation should continue and when it should stop. Because outside support is limited at sea, the weight of maintenance and safety judgment is unusually high.

### Social Media Manager

- URL: https://ai-job-risk.net/jobs/social-media-manager
- Slug: social-media-manager
- Industry: marketing.
- Current score: score 77, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace social media managers? Post drafts and hashtags automate, but holding brand tone and handling backlash stays human work.
- Editorial overview: Social media managers do far more than write posts. Their role is to decide what a brand should say, what kind of reactions it should aim for, and how it should respond to different voices, all while keeping the brand's position and business goals in mind. The work involves not only metrics, but also tone, backlash risk, community mood, and relationships with creators, which makes it far more complex than it may look on the surface.

### Social Worker

- URL: https://ai-job-risk.net/jobs/social-worker
- Slug: social-worker
- Industry: healthcare.
- Current score: score 20, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace social workers? Benefit lookups and application drafts speed up, but loosening the fear that keeps someone stuck does not.
- Editorial overview: Social workers do far more than explain systems. Their work is to organize situations where poverty, family relationships, employment, housing, caregiving, and continuing medical needs are all intertwined, then connect the person to support they can actually use in real life. They do more than listen. They bridge the gap between institutions and daily living.

### Sociologist

- URL: https://ai-job-risk.net/jobs/sociologist
- Slug: sociologist
- Industry: science.
- Current score: score 29, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace sociologists? Coding and summarization accelerate, but how a question is asked can change the result, and that judgment stays human.
- Editorial overview: Sociologists do much more than gather data and report averages. Their work is to explain human behavior in relation to institutions, regions, class, organizations, and culture rather than as isolated individual traits. Although the job can look like survey collection and interview aggregation, its real value lies in how social structures are framed and interpreted.

### Software Engineer

- URL: https://ai-job-risk.net/jobs/software-engineer
- Slug: software-engineer
- Industry: technology.
- Current score: score 79, weekly change +1, week 2026-09-02.
- Latest score explanation: This week brought another concrete signal of longer-running coding autonomy through OpenAI’s persistent agent work, which affects implementation, debugging, and maintenance tasks central to software engineering. Risk rises only modestly because the Hugging Face hack also highlighted how unreliable autonomous agents remain without strong human review and governance.
- Editorial description: Will AI replace software engineers? Routine code and test scaffolding are already faster, but architecture, production quality, and incidents remain.
- Editorial overview: Software engineers do a great deal more than write code. They design systems that software can keep running safely and sustainably over time. By linking requirements definition, design, implementation, review, incident response, and improvement, they take responsibility for the product’s overall quality and extensibility. Their value lies not in implementing one isolated feature, but in judging how the software should be built so it can evolve and be operated reliably in the future.

### Software Tester

- URL: https://ai-job-risk.net/jobs/software-tester
- Slug: software-tester
- Industry: technology.
- Current score: score 93, weekly change +1, week 2026-09-02.
- Latest score explanation: Autonomous agent improvements continue to favor software testing because test execution, bug triage, and coverage expansion are highly digital and repeatable. With coding agents becoming more persistent, software testing remains one of the jobs at risk from AI and rises slightly relative to the list.
- Editorial description: Will AI replace software testers? BLS projects about 10% growth in openings through 2034, partly because AI-generated code all needs vetting.
- Editorial overview: Software testers are often closer to the execution side of quality work than QA engineers. In practice, they interact directly with apps and systems to check whether behavior matches expectations and whether defects can be reproduced. Compared with QA engineers, who design overall quality strategy, software testers are usually closer to the front line of test execution and hands-on validation.

### Sound Engineer

- URL: https://ai-job-risk.net/jobs/sound-engineer
- Slug: sound-engineer
- Industry: entertainment.
- Current score: score 57, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace sound engineers? Noise cleanup and rough setups automate quickly, but live troubleshooting in a real room does not.
- Editorial overview: A sound engineer does more than clean up audio. The work includes judging recording conditions, balancing priorities in a live or studio environment, coordinating with performers and other departments, and creating sound that matches the intent of the production.

### Stock Trader

- URL: https://ai-job-risk.net/jobs/stock-trader
- Slug: stock-trader
- Industry: finance.
- Current score: score 57, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace stock traders? Speed-driven execution already belongs to algorithms; cutting exposure when a model breaks does not. See what remains.
- Editorial overview: Stock traders do far more than watch prices and place orders. Their job is to judge where to take risk and where to step back by reading supply and demand among market participants, the way news is being priced in, volatility, and position imbalances. Even in short-term trading, the work is a constant process of information processing and risk management.

### Supply Chain Analyst

- URL: https://ai-job-risk.net/jobs/supply-chain-analyst
- Slug: supply-chain-analyst
- Industry: logistics.
- Current score: score 57, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace supply chain analysts? Demand forecasts and KPI views automate, but explaining why a spike or shortage happened stays human.
- Editorial overview: Supply chain analysts do much more than compile numbers. They look across demand, inventory, procurement, transport, stockouts, and lead times to determine where bottlenecks exist and what changes would move the broader network closer to an optimum. Their responsibility is both reporting and turning analysis into insight that decision-makers can actually use.

### Supply Chain Manager

- URL: https://ai-job-risk.net/jobs/supply-chain-manager
- Slug: supply-chain-manager
- Industry: logistics.
- Current score: score 56, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace supply chain managers? Demand forecasts and scenario models automate; choosing what to protect during a shortage does not.
- Editorial overview: A supply chain manager does far more than oversee logistics. The role is about setting priorities so the entire supply network keeps moving, while balancing procurement, inventory, transportation, demand, warehousing, staffing, and supplier risk. The job is not only about day-to-day operations, but also about deciding what to protect when disruption hits.

### Surgeon

- URL: https://ai-job-risk.net/jobs/surgeon
- Slug: surgeon
- Industry: healthcare.
- Current score: score 10, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace surgeons? Preoperative planning and operative notes speed up, but deciding whether to operate and adapting mid-case do not.
- Editorial overview: Surgeons do much more than perform operations. Their work is to decide which intervention will actually benefit the patient by integrating imaging, tests, physical condition, comorbidities, surgical indication, procedure choice, and postoperative management. Their responsibility lies not only in technical skill, but also in the heavy judgment of whether surgery should be done at all.

### Surveying Technician

- URL: https://ai-job-risk.net/jobs/surveying-technician
- Slug: surveying-technician
- Industry: construction.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace surveying technicians? Drones, point clouds and draft drawings automate, but setting control points and judging error do not.
- Editorial overview: Surveying technicians do far more than collect coordinates. Their job is to establish accurate reference points for terrain, boundaries, structures, and construction positions so that design and construction can proceed without critical errors. Beyond operating equipment, they are responsible for where control points are set, how line of sight is secured, how error is managed, and how site constraints are handled.

### Sustainability Consultant

- URL: https://ai-job-risk.net/jobs/sustainability-consultant
- Slug: sustainability-consultant
- Industry: consulting.
- Current score: score 39, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace sustainability consultants? ESG data and disclosure drafts automate, but choosing which issues a business changes now does not.
- Editorial overview: A sustainability consultant does more than prepare disclosure materials. The role is about deciding what should be treated as a priority issue and how far it should be built into the business, while balancing regulatory requirements, customer pressure, business structure, operational data, and investor expectations. It is a job that requires drawing an implementable line rather than staying at the level of ideals.

### System Administrator

- URL: https://ai-job-risk.net/jobs/system-administrator
- Slug: system-administrator
- Industry: technology.
- Current score: score 72, weekly change +1, week 2026-09-02.
- Latest score explanation: Meta’s testing of robots to swap cables and reset servers in data centers is a direct deployment signal against portions of systems administration and routine infrastructure maintenance. Combined with stronger AI agent orchestration, the news slightly raises risk for admins handling repeatable monitoring and remediation work.
- Editorial description: Will AI replace system administrators? Runbooks and first-pass triage automate, but access decisions and change-impact calls stay human.
- Editorial overview: System administrators do far more than maintain servers. Their role is to keep internal and external systems running day after day by managing accounts, permissions, backups, patches, incident response, and operating procedures. Compared with flashy development work, the role places more weight on stable operation and accident prevention, but that quiet reliability directly affects the productivity of the entire organization.

### Tax Preparer

- URL: https://ai-job-risk.net/jobs/tax-preparer
- Slug: tax-preparer
- Industry: finance.
- Current score: score 70, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace tax preparers? Form entry and routine estimates automate, but confirming eligibility and catching evidence gaps still do not.
- Editorial overview: Tax preparers do far more than fil in tax forms. They gather required materials, support calculations, check whether eligibility conditions are actually met, work backward from filing deadlines, and organize facts and documentation so that a filing can be defended later. The role sits between numerical processing and compliance judgment.

### Taxi Driver

- URL: https://ai-job-risk.net/jobs/taxi-driver
- Slug: taxi-driver
- Industry: transportation.
- Current score: score 72, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace taxi drivers? Dispatch, demand forecasting and route calculation automate, but passenger reading and drop-off safety calls do not.
- Editorial overview: Taxi drivers do much more than carry passengers to a destination. They provide safe and comfortable travel tailored to each rider's circumstances, taking into account road conditions, time of day, neighborhood safety, luggage, and even passenger illness.

### Teacher

- URL: https://ai-job-risk.net/jobs/teacher
- Slug: teacher
- Industry: education.
- Current score: score 23, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace teachers? Lesson materials, test questions, and homework outlines generate in seconds; classroom judgment and trust do not.
- Editorial overview: Teachers do much more than explain subject knowledge. Their work includes planning lessons, identifying where students get stuck, managing the classroom, communicating with parents, and supporting each student’s understanding and growth. The value of the role comes not only from teaching the material, but also from guidance in daily life, assessment, career support, and building the atmosphere of a group.

### Teaching Assistant

- URL: https://ai-job-risk.net/jobs/teaching-assistant
- Slug: teaching-assistant
- Industry: education.
- Current score: score 37, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace teaching assistants? Initial grading and submission tracking automate; being the person students ask first still does not.
- Editorial overview: Teaching assistants do much more than help run a class. They stay close to the moments when students get stuck, notice confusion early, and support the day-to-day learning experience in ways instructors often cannot. Their value lies in helping keep the learning environment accessible, responsive, and manageable.

### Technical Writer

- URL: https://ai-job-risk.net/jobs/technical-writer
- Slug: technical-writer
- Industry: technology.
- Current score: score 68, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace technical writers? FAQ and procedure drafts are cheap now, but catching missed assumptions and version differences is not.
- Editorial overview: Technical writers turn product and system specifications into documents that users and operations teams can actually understand. Through help content, manuals, API documentation, user procedures, release notes, and FAQs, they connect the language of the people who build systems with the language of the people who use them. Their role is not merely to write explanations, but to communicate complex systems without creating misunderstanding.

### Telemarketer

- URL: https://ai-job-risk.net/jobs/telemarketer
- Slug: telemarketer
- Industry: marketing.
- Current score: score 83, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace telemarketers? BLS projects the occupation shrinking toward roughly 74,100 positions by 2034. See which calls AI voice agents take first.
- Editorial overview: Telemarketers make phone calls to introduce products or services, but the real substance of the role is deciding, in a very short interaction, whether the other party has any real interest and whether the conversation is worth continuing. The purpose may vary, from prospecting and re-proposing to reactivating dormant customers or setting appointments, but in every case the job depends on drawing out a meaningful reaction in a brief touchpoint.

### Therapist

- URL: https://ai-job-risk.net/jobs/therapist
- Slug: therapist
- Industry: healthcare.
- Current score: score 11, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace therapists? Hinge Health says its AI motion tracking cuts care-team hours per patient about 97 percent. See what still needs a person.
- Editorial overview: Therapists, physical, occupational, and rehabilitation therapists in particular, do a great deal more than administer pre-set exercises. Their work is to adjust the amount, order, and form of support to fit a person's physical condition, recovery stage, pain, anxiety, and life goals. The role involves not just delivering techniques but designing support that the person can actually continue at home.

### Tour Guide

- URL: https://ai-job-risk.net/jobs/tour-guide
- Slug: tour-guide
- Industry: hospitality.
- Current score: score 35, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace tour guides? Multilingual narration and route info automate, but reading a group's boredom and pacing the day does not.
- Editorial overview: A tour guide does more than provide tourism information. The role is to watch the mood of the participants and turn the experience of the place into time that feels meaningful. Explanation, movement management, atmosphere-building, and flexible response all come together in the job.

### Train Operator

- URL: https://ai-job-risk.net/jobs/train-operator
- Slug: train-operator
- Industry: transportation.
- Current score: score 66, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace train operators? Normal-section control and delay prediction automate; platform safety checks and restart calls stay human.
- Editorial overview: Train operators do far more than run a train at the prescribed speed. They keep rail operations safe by reading track conditions, platform safety, signals, delay propagation, and passenger movement. Their role is to balance punctuality with safety.

### Training Specialist

- URL: https://ai-job-risk.net/jobs/training-specialist
- Slug: training-specialist
- Industry: consulting.
- Current score: score 52, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace training specialists? Slides, quiz drafts and survey summaries automate, but designing for real behavior change does not.
- Editorial overview: A training specialist does more than create materials. The role is about designing learning so that it actually changes on-the-job behavior, while considering organizational issues, learner understanding, the behaviors needed in practice, and managerial support. The responsibility is both delivery and lasting adoption.

### Translator

- URL: https://ai-job-risk.net/jobs/translator
- Slug: translator
- Industry: media.
- Current score: score 74, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace translators? Machine translation already drafts business and news text, so value shifts to meaning, audience fit, and final QC.
- Editorial overview: Translators do much more than replace words from one language with words in another. Their job is to judge which translation will convey meaning without distortion while taking into account the intent of the original text, the background knowledge of the target reader, industry terminology, and cultural implications. In areas such as law, IT, medicine, and public communications, where even a small wording shift can cause major problems, a translator's judgment determines quality.

### Travel Agent

- URL: https://ai-job-risk.net/jobs/travel-agent
- Slug: travel-agent
- Industry: hospitality.
- Current score: score 71, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace travel agents? Itinerary drafts and price comparisons automate; building a plan that survives cancellations does not.
- Editorial overview: Travel agents do much more than book hotels and airline tickets. Their role is to build an itinerary that works as a whole while considering the purpose of the trip, budget, travel companions, length of stay, local transportation, safety, visa requirements, and cancellation risk. The level of judgment required differs greatly between leisure travel, corporate trips, and group travel, so the job cannot be reduced to simple search assistance.

### Truck Driver

- URL: https://ai-job-risk.net/jobs/truck-driver
- Slug: truck-driver
- Industry: transportation.
- Current score: score 77, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace truck drivers? Aurora ran driverless freight on about ten lanes in 2026, but loading and delivery sites stay human work.
- Editorial overview: Truck drivers do much more than move freight from one place to another. They make transport work by managing load condition, delivery-site constraints, road situations, and unloading arrangements. By 2026, driverless trucks from companies like Aurora Innovation and Kodiak AI are hauling real freight on selected highway corridors, but only on fixed lanes between two points, not the full door-to-door job.

### Tutor

- URL: https://ai-job-risk.net/jobs/tutor
- Slug: tutor
- Industry: education.
- Current score: score 34, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace tutors? Explanations, similar questions and study plans automate, but diagnosing why one student is stuck still needs a person.
- Editorial overview: Tutors do a great deal more than explain how to solve problems. They adjust learning pace and teaching style to each student’s level of understanding, personality, home environment, goals, and sense of weakness. Because one-on-one instruction is more personal than school lessons, the role often includes not only academic explanation, but also support for study habits and emotional stability.

### UI Designer

- URL: https://ai-job-risk.net/jobs/ui-designer
- Slug: ui-designer
- Industry: creative.
- Current score: score 64, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace UI designers? Wireframes and component drafts generate fast, but screen priority and exception states still need a designer.
- Editorial overview: A UI designer is more than someone who makes screens look nice. The role is about creating screens users can act through without hesitation by designing information hierarchy, visual clarity, and consistent component behavior. The responsibility covers both appearance and interaction.

### Urban Farmer

- URL: https://ai-job-risk.net/jobs/urban-farmer
- Slug: urban-farmer
- Industry: agriculture.
- Current score: score 43, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace urban farmers? Environmental control and repeated procedures automate; choosing what deserves the limited space does not.
- Editorial overview: An urban farmer does more than grow crops in a controlled environment. The role also includes deciding what kind of value can be sold in a city, adjusting cultivation and sales together, and operating within highly limited space and resources.

### Urban Planner

- URL: https://ai-job-risk.net/jobs/urban-planner
- Slug: urban-planner
- Industry: construction.
- Current score: score 38, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace urban planners? Land-use data, traffic drafts and zoning comparisons speed up, but defining what a place should become does not.
- Editorial overview: Urban planners do far more than draw maps. Their work is to design a region's future by looking at land use, transportation, disaster resilience, landscape, demographics, residents' daily lives, project feasibility, and administrative systems together. Unlike designing a single building, they must think about how to organize entire areas where many interests intersect.

### UX Designer

- URL: https://ai-job-risk.net/jobs/ux-designer
- Slug: ux-designer
- Industry: creative.
- Current score: score 41, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace UX designers? Personas, user flows and research summaries automate, but deciding which pain point actually matters does not.
- Editorial overview: A UX designer is more than someone who creates usable screens. The role is about understanding what users expect, where they get confused, and what kind of experience leads to continued use or understanding, then designing the entire experience structure. The responsibility is broader than a single screen and extends across the whole usage context.

### Veterinarian

- URL: https://ai-job-risk.net/jobs/veterinarian
- Slug: veterinarian
- Industry: healthcare.
- Current score: score 17, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace veterinarians? Imaging support and draft records speed up, but animals cannot describe symptoms and every owner's limits differ.
- Editorial overview: Veterinarians do much more than treate animal disease. Their work is to decide how far treatment should go by considering symptoms, behavior changes, test results, the living environment, the owner's judgment, and cost constraints together. Because the patient cannot explain their condition in words, observation and owner-provided information carry especially heavy importance.

### Veterinary Assistant

- URL: https://ai-job-risk.net/jobs/veterinary-assistant
- Slug: veterinary-assistant
- Industry: healthcare.
- Current score: score 40, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace veterinary assistants? Booking and record entry automate first, but safe restraint and spotting subtle changes stay human.
- Editorial overview: Veterinary assistants do a great deal more than handle miscellaneous clinic tasks. Their work supports a veterinary hospital's safe and smooth operation through restraint, cleaning, supply preparation, clinical assistance, observation of hospitalized animals, and owner support. They work close to both animals and people, helping hold the clinical environment together.

### Video Editor

- URL: https://ai-job-risk.net/jobs/video-editor
- Slug: video-editor
- Industry: media.
- Current score: score 54, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace video editors? Transcription, silence cuts and subtitle drafts automate; deciding what to cut and what viewers feel does not.
- Editorial overview: Video editors do much more than connect clips. They design the flow of time itself. By deciding which cuts to keep, where to create breathing room, and when to show which information, they can create completely different impressions from the same raw footage. The work is not only technical operation, but the translation of intent into presentation.

### Waiter

- URL: https://ai-job-risk.net/jobs/waiter
- Slug: waiter
- Industry: hospitality.
- Current score: score 31, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace waiters? Order entry, payment and basic menu guidance automate, but service timing and reading the room stay human work.
- Editorial overview: A waiter does more than carry food. The role is to keep the guest's dining experience moving comfortably by coordinating order-taking, delivery timing, conversational distance, and the atmosphere at the table. It directly shapes the impression of the restaurant.

### Warehouse Manager

- URL: https://ai-job-risk.net/jobs/warehouse-manager
- Slug: warehouse-manager
- Industry: logistics.
- Current score: score 49, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace warehouse managers? Picking optimization and staffing simulations automate, but blocked aisles and safety trade-offs need a person.
- Editorial overview: Warehouse managers do far more than overse inventory. They create conditions where inbound receiving, put-away, picking, shipping, staffing, safety, and storage quality can keep moving without blockage across the site. Their responsibility involves more than monitoring numbers; it also involves spoting congestion and accident risk early enough to keep the floor running.

### Warehouse Operator

- URL: https://ai-job-risk.net/jobs/warehouse-operator
- Slug: warehouse-operator
- Industry: logistics.
- Current score: score 65, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace warehouse operators? Picking instructions and scan checks automate; unreadable labels and collapsed packaging still need people.
- Editorial overview: Warehouse operators do much more than move goods. Through receiving, put-away, picking, packing, and loading, they connect the flow accurately while responding to volume swings and item-specific handling needs. Their responsibility is both speed and preventing mis-shipments and damage while work continues moving.

### Waste Management Specialist

- URL: https://ai-job-risk.net/jobs/waste-management-specialist
- Slug: waste-management-specialist
- Industry: environment.
- Current score: score 37, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace waste management specialists? Forms and volume aggregation automate, but stopping sorting drift on a live floor does not.
- Editorial overview: Waste management specialists do a great deal more than dispos of waste. Their role is to design the operation: classifying waste, planning storage methods and collection routes, handling legal compliance, and managing contractors. They need to understand both floor workflow and regulation in order to prevent accidents and violations.

### Water Treatment Operator

- URL: https://ai-job-risk.net/jobs/water-treatment-operator
- Slug: water-treatment-operator
- Industry: environment.
- Current score: score 41, weekly change -1, week 2026-09-02.
- Latest score explanation: Reports of hackers targeting more than 100 US water systems and broader concern about AI-enabled cyber incidents make human oversight more valuable in this infrastructure role. Because the job involves safety-critical monitoring and response, the week’s news slightly lowers near-term replacement risk.
- Editorial description: Will AI replace water treatment operators? Threshold monitoring automates, but trouble often shows in sight and sound before it shows in the numbers.
- Editorial overview: Water treatment operators do far more than switch equipment on and off. Their role is to maintain stable, safe supply or discharge treatment by watching flow rate, water quality, chemical dosing, and equipment condition. They support critical infrastructure through both screen-based monitoring and physical rounds on site.

### Web Developer

- URL: https://ai-job-risk.net/jobs/web-developer
- Slug: web-developer
- Industry: technology.
- Current score: score 67, weekly change +1, week 2026-09-02.
- Latest score explanation: Persistent coding agents and continued enterprise experimentation with autonomous software workflows increase pressure on template-based site building, debugging, and routine front-end updates. The rise is limited because client communication, integration complexity, and review still slow full replacement.
- Editorial description: Will AI replace web developers? AI mass-produces plausible screens, but browser differences, performance, SEO, and accessibility still break them.
- Editorial overview: Web developers do far more than make screens look good or wire up the back end. Their job is to make websites and web apps work properly in the browser while handling UI, API integration, performance, accessibility, and ease of ongoing operation. A more accurate description is that they use technology to create an experience people can use without getting lost.

### Welder

- URL: https://ai-job-risk.net/jobs/welder
- Slug: welder
- Industry: manufacturing.
- Current score: score 33, weekly change +0, week 2026-09-02.
- Editorial description: Will AI replace welders? Parameter settings and inspection records automate first, but heat distortion and misaligned tacks stay human calls.
- Editorial overview: Welders do much more than join pieces of metal together. Their job is to make both strength and workability hold up by accounting for material type, thickness, working position, heat input, distortion, and the needs of later processes. It is not enough to run a bead according to the drawing; welders are also responsible for adjusting conditions to fit the real workpiece in front of them.

## Industry Editorial Context

### Agriculture

- URL: https://ai-job-risk.net/industries/agriculture
- Lead: Agriculture runs on decisions made in a specific field, on a specific day, under weather that never repeats exactly the same way twice. Yield monitors, soil sensors, and satellite imagery now feed data back faster than any agronomist could walk a row, and that data genuinely changes how seed rates, irrigation schedules, and spraying windows get planned across a season. But a forecast is not a harvest. Hail, a broken irrigation pivot, a fungus that shows up two days early, or a buyer who suddenly changes a delivery spec still forces someone in boots to reread the situation and act, which keeps this industry from becoming a pure data problem.
- Overview: Read this industry by separating the desk-and-dashboard side of farm management from the work that only makes sense standing in the field or the barn. Prescription maps, commodity price tracking, compliance recordkeeping, and equipment maintenance scheduling are increasingly software-driven and respond well to automation support. Herd health checks, irrigation timing under an unexpected heat wave, harvest-window calls made against an incoming storm, and machinery repair when something breaks mid-season all depend on a person reading conditions that sensors only partially capture, so those tasks resist the same pace of change.
- Automation: AI and precision-agriculture tools move first into variable-rate seeding and fertilizer prescription maps, yield-monitor data logging, drone and satellite imagery that flags crop stress, automated steering on tractors and combines, and grain-marketing analytics that suggest when to sell into the futures market. Livestock operations increasingly use ear-tag sensors and barn camera systems to flag a sick or lame animal before a handler would otherwise notice the change. Automation stalls on diagnosing a new pest or disease pattern the model has not seen before, deciding whether to spray given a shifting weather window, handling calving and lambing complications, and repairing hydraulics or onboard electronics on aging equipment out in the field with no cell signal.
- Human: The roles that stay durably human are the ones that combine mechanical aptitude with field judgment: equipment operators who can improvise a repair with whatever is in the shop, herd managers who read subtle behavior changes in livestock long before a sensor threshold trips, agronomists who walk a field to confirm what a satellite image only suggests from a distance, and farm managers who decide what to plant, when to sell, and when to cut losses on a failing crop under real financial pressure. These roles carry consequences that a bad model output simply does not.
- Reading: When you look at scores in this industry, separate roles built mostly around monitoring dashboards and compliance paperwork from roles built around handling livestock, operating and repairing machinery, or making real-time calls about weather and crop condition in the field. The former absorbs AI support quickly, while the latter keeps a person in the loop because the cost of a wrong call, a sick herd, or a lost harvest window is high, immediate, and often impossible to reverse once it happens.

### Construction

- URL: https://ai-job-risk.net/industries/construction
- Lead: Construction produces enormous amounts of documentation, drawings, and schedules, and AI tools now compare revisions, flag clashes in a Building Information Model, and generate quantity takeoffs faster than an estimator working through drawings alone. That is real and genuinely useful. But no software has to stand on a slab and decide whether the rebar was poured wrong, or tell a crew to stop pouring when the ground turns out different from what the geotechnical report predicted. The industry's risk profile is genuinely split between office work that automates readily and site work that does not.
- Overview: The clearest way to read this industry is to separate preconstruction and back-office work from work that happens on an active, physically changing site. Bid preparation, scheduling software, drawing markup, cost estimating, and progress reporting increasingly run through AI-assisted tools that speed up document-heavy tasks across an entire project lifecycle. Site supervision, trade coordination, safety judgment, and the constant renegotiation that happens when a delivery is late or an inspection fails remain grounded in a person who is physically present on-site and personally accountable for what actually gets built.
- Automation: AI tools move fastest through BIM clash detection across mechanical, electrical, and structural drawings, automated quantity takeoffs, AI-assisted scheduling that models trade sequencing across a project timeline, drone-based progress photography compared against design plans, and safety-compliance checklists generated from inspection photos. It stalls where site reality diverges from the model: unexpected soil or buried-utility conditions, a subcontractor who falls behind schedule and forces the whole sequence to be replanned, an inspector who fails a concrete pour for a reason no checklist captured, and safety calls made in the moment when a crane lift, trench, or scaffold looks wrong to an experienced eye.
- Human: What remains durably human are the site superintendent who resequences three trades after a delivery delay, the safety officer who stops work on instinct before an incident happens rather than after, the foreman who catches that a wall is out of plumb before it gets closed in behind drywall, and the project manager who renegotiates scope with a client and subcontractors when conditions do not match the drawings at all. Electricians, plumbers, and ironworkers who adapt fixed designs to an irregular structure hold similar value. These roles carry liability and physical risk that a scheduling tool never answers for.
- Reading: Score this industry by asking whether a role lives mostly in drawings, spreadsheets, and reports, or mostly on an active site managing trades, safety, and physical conditions that change every day the project runs. Estimating and project-controls roles trend toward faster automation as document tools mature. Superintendents, safety leads, and skilled trades who adapt to what the site actually gives them keep more of their weight in the overall risk score.

### Consulting

- URL: https://ai-job-risk.net/industries/consulting
- Lead: Consulting sells synthesis: turning research, benchmarking data, and client interviews into a recommendation someone will actually act on. Large parts of that pipeline are already faster with AI, from pulling comparable-company data to drafting a first slide deck overnight. That has genuinely changed how junior consultants spend their weeks. But the value a client actually pays for is framing the right problem and standing behind a recommendation once it meets organizational politics, and that part of the job has not gotten easier to automate, because it depends on reading one specific client's context, not on producing more analysis faster.
- Overview: Read consulting work by separating research and production from framing and judgment, since AI reshapes them at very different rates. Market sizing, competitor benchmarking, literature synthesis, and first-draft deck production are information-processing tasks that move quickly with AI support once the source material is available. Diagnosing what problem a client actually has, choosing which recommendation will survive internal politics, and building the credibility to be believed in the room are advisory tasks that depend on context AI does not have access to, and that gap explains most of the variation in how this industry is affected.
- Automation: AI moves first through desk research, competitor and market benchmarking, transcript and interview synthesis, first-pass slide drafting, and financial modeling templates that used to consume a junior analyst's first week on a project. Building comparable-company sets or summarizing a stack of industry reports overnight is now routine work handled largely by machine. It stalls on client-specific diagnosis: figuring out why a reorganization keeps failing at one particular company despite working elsewhere, negotiating scope with a skeptical project sponsor, and presenting a recommendation to a room that already has strong, entrenched opinions about what the answer should be before the meeting even starts.
- Human: What holds up in consulting is the ability to frame an ambiguous problem correctly and to carry a recommendation through an organization's internal politics. Engagement managers who read what a client actually needs versus what they officially asked for, partners who have the relationship capital to be believed even when the message is unwelcome, and specialists who can defend a recommendation under hostile questioning from a skeptical executive are doing work that depends on trust built case by case over years, not on faster research turnaround or a cleaner slide.
- Reading: Score a consulting role by asking whether it is mostly research production or mostly client-facing judgment under pressure. Analyst-heavy roles built around benchmarking, synthesis, and deck production show higher exposure because that output is increasingly commoditized. Roles centered on problem framing, client relationships, and recommendations that have to survive contact with a real, political organization score lower, since faster research does not by itself produce a recommendation anyone will actually implement.

### Creative

- URL: https://ai-job-risk.net/industries/creative
- Lead: Creative and design work is exposed to AI in an unusually direct way: image generation, layout drafting, and rapid variation are exactly what these tools were built to do, and a designer who once spent a day producing three concepts can now generate dozens in minutes. Agencies and in-house design teams already use generative tools for mockups, mood boards, and first-pass layouts. But a client is not paying for volume of options. They are paying for someone who understands the brief well enough to know which option actually solves the problem, and who can defend that choice when it's challenged.
- Overview: Separate creative work that generates options from work that sets the standard those options are judged against. Producing layout variants, generating mockups, drafting visual directions, and building asset variations for different formats is generation: AI already does large volumes of this quickly. Interpreting a client's brief, deciding which direction solves the underlying problem, and maintaining a consistent creative standard across a body of work is direction: it requires understanding intent never fully written into the brief. A design team can generate far more options than before while still needing the same small number of people who decide which option is right.
- Automation: AI moves first into mockup generation, layout variation, image and asset generation from a brief, typography and color-palette exploration, and producing multiple format adaptations of an approved design for different channels. Stock-image and template generation that used to require a photo shoot or illustrator now happens in minutes. It stalls on translating an ambiguous client brief into an actual creative direction, making the trade-off call between competing good options, defending a creative choice to a client who pushes back, and maintaining a coherent visual identity across a large body of work over time rather than producing one good asset in isolation.
- Human: What stays durably human is setting the creative standard and translating intent into a defensible direction. Creative directors who choose which concept goes to the client and why, art directors who keep a brand's visual identity coherent across hundreds of assets and multiple contributors, and senior designers who can explain the reasoning behind a layout choice under client pushback carry work that generation tools do not replace. The judgment of knowing when a technically competent option is still the wrong option for this brief and this client depends on experience that isn't captured in a prompt.
- Reading: Read creative scores by separating generation from direction. A junior designer or production artist producing variations and format adaptations scores higher risk because that output is now easy to generate in volume. A creative director or senior art director scores lower because the job is choosing and defending direction, not producing more options. The size of a portfolio a person can generate matters far less to their risk score than how much of their job is deciding what belongs in it.

### Education

- URL: https://ai-job-risk.net/industries/education
- Lead: Education is built on explanation, and AI is very good at generating explanations, which is why it is landing here faster than in most people-facing fields. Lesson materials, practice questions, first drafts of feedback on writing, and administrative scheduling can now be produced in seconds by tools that used to take a teacher an evening. But teaching a room full of specific students is not the same task as generating content for an average one. The tension is between what a model can draft in general and what a teacher must do in particular: noticing that one student is disengaged, managing a classroom in the moment, and motivating someone who has already decided they are bad at the subject.
- Overview: Read AI's effect in education by separating content creation from the relationship of teaching. Generating a worksheet, drafting rubric-aligned feedback, translating a passage into simpler language, or building a first-pass lesson plan is now fast and largely automatable. Reading a specific student's confusion in real time, adjusting mid-lesson when the room isn't following, and rebuilding a discouraged learner's confidence are not compressed by the same tools, because they depend on a live read of one person or one group, not a generic pattern.
- Automation: AI moves first through content generation — quizzes, slide decks, differentiated reading passages — through first-pass grading of objective and short-answer work, through drafting tutoring explanations and practice problems, and through administrative scheduling, attendance, and routine parent communication. Learning-management systems already automate reminders and progress reports. It stalls where a real student is in the room: recognizing that a correct-looking answer hides a misunderstanding, managing behavior and attention during a lesson, adapting pace for a learner who is stuck, and holding a conversation with a struggling or upset student that a generic response would mishandle.
- Human: What stays durably human in education is reading people, not content. Classroom teachers managing a room's attention and behavior in real time, counselors and special-education staff working with a specific child's needs, coaches and mentors building motivation over a school year, and any educator diagnosing why a particular student is stuck rather than what the general topic requires — these depend on sustained, situational attention to one person that a model cannot replicate from a transcript.
- Reading: For education roles, weigh how much of the work is producing generic materials or grading standardized responses versus reading and motivating specific learners in real time. Roles centered on content drafting, routine grading, or scheduling should expect a higher score, reflecting how much of that output can already be generated. Roles centered on classroom management, individualized support, or building a struggling student's confidence should read a lower score as a genuine reflection of what still requires a present, attentive adult.

### Energy

- URL: https://ai-job-risk.net/industries/energy
- Lead: Energy production and distribution generate huge volumes of sensor data, from turbine vibration readings to grid load curves updated every few seconds, and AI-driven predictive maintenance and load forecasting are now standard practice in control rooms. That data advantage is real and measurable. But a transformer fire, a gas leak, or a sudden grid frequency swing does not wait for a model to finish retraining, and the operators who intervene during those events carry personal and public safety responsibility that no dashboard can absorb on its own behalf.
- Overview: This industry splits cleanly between control-room analytics and field-and-plant operations that respond to physical failure in real time. Load forecasting, predictive maintenance scheduling, SCADA data analysis, and regulatory emissions reporting are increasingly automated and improving quickly across utilities and producers alike. Emergency switching, live-line electrical work, outage restoration after storm damage, and the judgment calls made during an abnormal grid event or a well-control incident stay with people trained to act under pressure with incomplete information and real safety stakes.
- Automation: AI moves first into predictive maintenance models that flag failing transformers or turbine bearings before they actually fail, load-forecasting systems that balance generation against demand across the grid, SCADA anomaly detection, drone inspection of pipelines and transmission towers, and automated compliance reporting for emissions and safety regulators. It stalls on live-line electrical work performed on energized equipment, emergency response to a gas leak or well blowout, black-start recovery after a wide-area outage, and any situation where a lineworker or plant operator must decide in real time whether a reading reflects a sensor fault or an actual dangerous condition on the ground.
- Human: Durable roles include lineworkers who restore power after storm damage in conditions no drone can fully assess from the air, control-room operators who override automated dispatch during a live grid emergency, plant operators who shut down equipment on judgment before an automated alarm even confirms failure, and safety-critical field crews performing hot work on energized lines and pressurized systems. Well-control specialists on drilling and production sites carry a similar weight of judgment. These roles carry direct physical risk and public-safety accountability that stays attached to a named, licensed person.
- Reading: Read scores here by separating control-room and back-office analytics from field operations and emergency response work. Forecasting analysts and regulatory reporting staff face faster automation pressure as tools mature and adoption spreads across the sector. Field technicians, plant operators, and emergency-response crews keep more human weight in the score because the cost of a wrong call includes safety incidents and widespread outages, not simply inefficiency or slower reporting.

### Entertainment

- URL: https://ai-job-risk.net/industries/entertainment
- Lead: Entertainment is where AI's ability to generate rough footage, music beds, dialogue passes, and visual effects assets collides directly with a business whose entire value proposition is holding an audience's attention. Editing assists that once took a post-production team days can now produce a workable cut in a fraction of the time, and tools that generate temp music, storyboards, or crowd shots are already used on real productions. But none of that decides whether the story is worth telling or whether an audience will feel anything watching it. That judgment, and the performance that carries it, still sits with people.
- Overview: Separate entertainment work that produces assets from work that decides what belongs in front of an audience. Rough-cut editing, temp scoring, dialogue cleanup, background and crowd generation, and visual-effects prep are asset production: AI already speeds these up substantially on real sets and in real post houses. Deciding what story to greenlight, how a scene should be paced for emotional effect, and what a performance needs to convey are audience-facing judgment: they depend on taste and read of human reaction that a generation model does not have. A production can lean hard on the first category while the second stays almost entirely human.
- Automation: AI moves first into rough-cut assembly, temp music and sound design passes, dialogue and audio cleanup, background and crowd-shot generation, color-grading first passes, storyboard and previsualization drafts, and localization such as dubbing scripts or subtitle generation. Visual-effects studios already use generative tools for rotoscoping, matte cleanup, and de-aging passes that used to require large teams. It stalls on greenlighting decisions, directing a performance to hit a specific emotional beat, editing for pacing that serves the story rather than just assembling footage, and the live, unscripted judgment a performer or showrunner exercises when something isn't working and needs to change on the spot.
- Human: What stays durably human is performance, direction, and the judgment about what will actually move an audience. Directors who shape a scene's pacing and tone, actors who bring a script to life in ways no draft anticipated, and showrunners who decide where a season's story goes next carry work that resists generation. Editors who make the final call on which take serves the story, and live performers whose value is presence in the room, depend on a kind of read of human reaction that stays stubbornly human even as the tools around them get faster.
- Reading: Read entertainment scores by separating production-support roles from performance and creative-direction roles. A visual-effects technician, colorist, or post-production assistant scores higher risk because the work is asset generation that tools already do well. A director, showrunner, or lead performer scores lower because the job is deciding what will move an audience, which is not a task a model can be handed. The same production can carry both high-risk and low-risk roles side by side.

### Environment

- URL: https://ai-job-risk.net/industries/environment
- Lead: Environmental work depends on satellite data, sensor networks, and modeling software that AI now accelerates significantly, from tracking deforestation across large regions to flagging water-quality anomalies in real time. That analytical layer is genuinely faster than it used to be even a few years ago. But a contaminated site, a contested permit, or a community meeting about a proposed landfill still requires someone who can walk the ground, interpret ambiguous readings, and explain a defensible decision to regulators, companies, and residents who each want a different answer.
- Overview: Separate the remote-sensing and modeling side of environmental work from the site-assessment and stakeholder side of the job. Satellite monitoring, pollution-dispersion modeling, compliance-report drafting, and permit-application review are increasingly AI-assisted and move faster each year as datasets and tools improve across agencies and consultancies. Site inspections, contamination sampling, interpreting conflicting data from a failing sensor network, and mediating between regulators, companies, and affected communities all depend on judgment and earned trust that a model simply cannot supply on its own.
- Automation: AI tools move first into satellite and drone monitoring of land use, deforestation, and emissions; water and air quality anomaly detection drawn from sensor networks; automated drafting of environmental impact statements and permit applications; and pattern analysis across large regulatory compliance datasets spanning many facilities at once. It stalls when a sensor reading looks wrong and someone has to physically travel to verify it, when a contamination plume behaves unlike the model predicted, and when a decision needs a defensible, publicly explainable judgment call about acceptable risk at one particular site with real people living and working nearby.
- Human: Roles that stay durably human include field inspectors who take soil and water samples and catch what a sensor network misses entirely, remediation specialists who adjust cleanup plans when a site does not behave as the model predicted, and regulatory liaisons who explain risk tradeoffs to communities, companies, and agencies that each carry competing interests. Environmental compliance officers who must sign off on a facility's actual conduct, not just its reported numbers, carry similar weight. These roles carry legal and reputational accountability that outlasts the modeling layer by a wide margin.
- Reading: Read this industry by asking whether a role is mostly building or reading models and reports, or mostly verifying conditions on-site and negotiating an outcome that stakeholders will actually accept. Data-analysis and reporting roles trend toward faster automation as tooling improves. Field inspection, remediation, and community- or regulator-facing roles keep more weight in the score because they carry direct accountability for real-world outcomes on the ground.

### Finance

- URL: https://ai-job-risk.net/industries/finance
- Lead: Finance runs on numbers that have to reconcile, documents that have to match, and reports that have to close on a fixed schedule, which is exactly the kind of structured, rule-bound work AI tools already handle well. Reconciliation, variance analysis, and first-pass credit or fraud screening are noticeably faster with model support than a few years ago. But finance is also a trust system built around accountability: someone has to sign off on a loan, defend a valuation to an auditor, or explain to a client why a portfolio lost money. That accountability, not the arithmetic behind it, is what keeps a specific person in the loop.
- Overview: Read a finance role by splitting it into the numbers it produces and the calls it makes on top of them. Bookkeeping, reconciliation, regulatory reporting, and routine ratio analysis are largely mechanical, follow well-defined rules, and move fast under automation once the data feeds are in place. Underwriting, audit sign-off, portfolio decisions, and client-facing advice involve judgment about ambiguous facts, incomplete information, and consequences that fall on a named person, so they resist compression even as the underlying data work speeds up considerably. The gap between these two categories is what the score is really measuring.
- Automation: AI moves first through bank and ledger reconciliation, invoice matching, expense processing, month-end close checklists, and first-pass variance reports that used to take an analyst days to assemble. Fraud-detection models already flag anomalous transactions faster than manual review, and credit-scoring systems produce an initial risk read before a human loan officer opens the file. Robotic process automation handles data entry between systems that were never built to talk to each other. It stalls on exception cases: a borrower with an unusual income pattern, a transaction that trips a fraud model but turns out legitimate, an audit finding that needs professional interpretation rather than a rule lookup, or a client whose situation doesn't fit the standard product.
- Human: What stays durably human in finance is accepting accountability for a judgment call that could be wrong. A credit officer who approves a loan the model flagged as borderline, an auditor who signs an opinion tied to their professional license, a financial advisor who talks a client out of a bad decision during a market panic, and a controller who explains a discrepancy to regulators are all doing work that requires standing behind a conclusion, not just producing one. Relationship-heavy roles like private banking and complex deal structuring depend on trust built over years, which a faster model does not replicate.
- Reading: When you look at a finance role's score, ask how much of the job is processing transactions versus owning a decision that could be costly if it turns out wrong. Back-office reconciliation, bookkeeping, and standard reporting roles tend to score higher on exposure because the work is repeatable and well-documented. Roles built around underwriting judgment, audit opinions, regulatory interpretation, or client trust score lower, even though they use much of the same software and data as the roles being automated around them.

### Government

- URL: https://ai-job-risk.net/industries/government
- Lead: Government work is unusually document-heavy: permit applications, benefits eligibility files, regulatory filings, FOIA requests, and case records move through the same structured paperwork that AI tools are good at processing at scale. That creates real efficiency opportunity in agencies long paper-bound and understaffed relative to caseload. But government decisions carry a constraint most private-sector work does not: a denied benefits claim, a contested zoning ruling, or a policy choice has to be explainable to the public and defensible on appeal, which keeps a named official accountable even when a system produced the recommendation.
- Overview: Separate government work into records processing and public-facing judgment, because the two move at very different speeds under automation. Intake of applications, document classification, records requests, and case-file preparation are procedural tasks that speed up substantially once digitized. Eligibility determinations that involve discretion, enforcement decisions, policy tradeoffs, and anything that ends up in a public hearing or a legal appeal involve interpreting rules against messy individual circumstances, and that interpretive work is where human judgment stays load-bearing regardless of how fast the paperwork moves.
- Automation: AI moves first through records digitization, FOIA and public-records search, benefits and permit intake triage, form processing, and drafting routine correspondence and internal reports. Case-management systems already pre-sort applications and flag missing documentation before a caseworker ever touches the file, and chatbots handle a growing share of routine constituent inquiries about status and requirements. It stalls where a decision has to survive an appeal, a hearing, or a records request: an eligibility case with unusual family or income circumstances, an enforcement action a business intends to contest, a zoning variance drawing neighborhood objections, or any policy choice that elected officials must defend publicly to constituents and the press.
- Human: Durable human roles in government center on public accountability: caseworkers who exercise discretion in genuine hardship situations, inspectors who make judgment calls on-site that a checklist doesn't fully capture, policy staff who weigh competing constituencies against each other, and officials who have to explain a decision at a public hearing or to a reporter. Procurement officers negotiating contract terms with vendors and program managers coordinating across departments with conflicting priorities also depend on institutional relationships and accumulated judgment that a records system doesn't carry.
- Reading: For government roles, weigh how much of the job is processing standardized applications versus making determinations that could be appealed or scrutinized publicly. Clerical and records-intensive roles tend to show higher exposure because the underlying documents are structured and repetitive. Roles involving discretion in individual cases, enforcement judgment, or public accountability score lower, because the requirement to justify a decision to the public and to survive an appeal is a constraint automation does not remove.

### Healthcare

- URL: https://ai-job-risk.net/industries/healthcare
- Lead: Healthcare runs on both information and touch, and AI is reshaping the information half faster than most staff expected. Clinical documentation, intake paperwork, and the endless back-and-forth with insurers consume hours that clinicians would rather spend with patients, and ambient scribing tools, coding assistants, and scheduling systems are already absorbing pieces of that load. The tension is that a hospital or clinic is not a records office. The moments that matter most — telling a patient what a diagnosis means, deciding on a treatment path, taking responsibility when something goes wrong — still require a licensed person who can be held accountable, examine a body, and sit with someone who is frightened.
- Overview: Read AI's effect in healthcare by separating paperwork from patient care. Anything that involves converting speech or data into a structured record — visit notes, billing codes, prior-authorization forms, appointment reminders — is being compressed by software that listens, drafts, and files. Anything that involves touching a patient, weighing an ambiguous set of symptoms, or delivering difficult news moves at the same pace it always has. A role heavy on the first kind of work looks exposed; a role built around the second holds steady even as the tools around it change.
- Automation: AI moves first through ambient documentation that turns a visit into draft clinical notes, through medical coding and billing software that suggests codes from a chart, through imaging triage tools that flag likely-abnormal scans for radiologist review, and through scheduling and intake systems that handle referrals and reminders. Pharmacy systems already flag drug interactions automatically. It stalls hard at the bedside: physical examination, interpreting an ambiguous or atypical presentation, choosing among treatment options with real tradeoffs, obtaining informed consent, and any moment where a patient needs to be reassured by a person who will answer for the outcome.
- Human: What stays durably human in healthcare is anything requiring hands, judgment under uncertainty, and accountability. Nurses assessing a patient who doesn't match the textbook picture, physicians weighing competing risks in a treatment decision, surgeons and proceduralists performing physical interventions, and therapists building trust over many sessions are not substitutable by a model. So is the work of breaking bad news, coordinating care across a confused or frightened family, and the legal and ethical responsibility that follows a clinician's signature on a chart.
- Reading: For healthcare roles, weigh how much of the job is documentation, coding, or scheduling versus hands-on assessment and decision-making under uncertainty. Roles concentrated in transcription, billing, or appointment logistics should read a higher score as a signal that routine throughput is already being absorbed by software. Roles centered on physical care, diagnosis in ambiguous cases, or consent and accountability should treat a lower score as reflecting genuine insulation, not an oversight in the model.

### Hospitality

- URL: https://ai-job-risk.net/industries/hospitality
- Lead: Hospitality sells an experience, but most of the transactions behind that experience are routine and data-heavy, which is exactly what AI automates well. Booking engines, dynamic pricing, ordering kiosks, and review-response drafting already run with little human input at many properties and restaurants. The tension is that guests remember the moments a system can't script: a host who notices a couple is celebrating an anniversary, a server who catches a problem before a guest complains, a manager who resolves a bad night in person. Automating the transaction layer doesn't touch the service layer that actually earns loyalty and tips.
- Overview: Read AI's effect in hospitality by separating the transaction from the encounter. Reservations, table and room assignments, inventory and ordering, dynamic pricing, and first-draft responses to online reviews are now handled largely by software. Reading a guest's mood, recovering a service failure face to face, and creating the small unscripted moments that make a stay or meal memorable are not, because they depend on live human presence and improvisation that a booking system or chatbot cannot supply.
- Automation: AI moves first through reservation and booking platforms, dynamic pricing engines that adjust room and menu prices to demand, self-service ordering kiosks and apps, inventory and supply forecasting, and drafted responses to online reviews. Back-office scheduling of shifts increasingly runs on automated forecasting of expected demand. It stalls on the floor: greeting and reading a guest in person, recovering a service failure with empathy and quick judgment, adapting service style to a specific table or room, and any moment where hospitality means noticing something a system was never told to look for.
- Human: What stays durably human in hospitality is presence. Front-desk staff and hosts managing a real guest's mood and unexpected requests, servers and bartenders reading a table and adjusting service on the fly, housekeeping and maintenance staff solving physical problems on-site, and managers stepping in to recover a guest's bad experience face to face all depend on being physically present and improvising in the moment, which is precisely what booking and ordering software cannot do.
- Reading: For hospitality roles, weigh how much of the job is processing bookings, orders, and standard requests versus delivering in-person service and recovering problems on the spot. Roles concentrated in reservations, ordering systems, or review-response drafting should expect a higher score, since that throughput is already largely automated. Roles built around face-to-face guest interaction, service recovery, or on-site problem-solving should read a lower score as reflecting real, durable value in being present.

### Legal

- URL: https://ai-job-risk.net/industries/legal
- Lead: Legal work sits at the center of the AI debate because so much of it runs on text: contracts, filings, case law, disclosure documents, and correspondence. Tools that search, summarize, and draft at scale genuinely change how legal teams spend their hours. But a law practice is not a document-processing pipeline. The parts that carry the most risk — deciding what a clause actually means, choosing a litigation strategy, advising a client who has to live with the outcome — still rest on human interpretation and on professional accountability that courts and regulators attach to a named person.
- Overview: The clearest way to read AI's effect on legal work is to separate roles built mainly on producing and moving documents from roles built on judgment, advocacy, and responsibility. Contract review, legal research, e-discovery, and first-draft generation are all areas where AI assistants now do in minutes what once took paralegals and junior associates hours, and that pressure falls hardest on high-volume, repeatable work. It falls far more lightly on the parts of the job where a lawyer has to commit to a position, weigh consequences, and stand behind advice — because speed of drafting is not the same as ownership of a decision.
- Automation: AI moves first and fastest through document review, precedent search, due-diligence checklists, contract markup, transcript and deposition summarization, and the first pass of routine drafting. E-discovery platforms already triage huge document sets, and contract-analysis models flag missing clauses and non-standard terms faster than manual review. It stalls on work that has to survive scrutiny: framing the actual legal question, judging which risks are acceptable, negotiating terms with a counterparty, and explaining a position to a judge, a regulator, or a board that will act on it. Those tasks depend on context, strategy, and a professional who can be held responsible if the reasoning fails.
- Human: The most durable human value in legal work is not knowing the law in the abstract — models can retrieve statutes and cases — but deciding how to use it for a specific client in a specific situation and being accountable for that decision. Litigators, negotiators, in-house counsel advising the business, and specialists handling novel or contested matters keep far more of their value than roles centered on high-volume document handling. Client trust, courtroom judgment, and the willingness to sign your name to advice are things a tool cannot assume on a lawyer's behalf.
- Reading: Read the score with the shape of the actual role in mind, not the label 'legal.' A job that is mostly search, review, and standardized drafting will look — and genuinely be — more exposed than one built on interpretation, negotiation, and responsibility, even though both sit under the same profession. The most honest reading is to ask how much of a given role is producing documents versus owning decisions that someone will act on and that can be traced back to a named person.

### Logistics

- URL: https://ai-job-risk.net/industries/logistics
- Lead: Logistics is one of the most heavily instrumented industries in the economy: route optimization software, warehouse robotics, and demand-forecasting models already run much of the daily plan before a human ever sees it. That plan looks efficient on a screen. It falls apart the moment a container is delayed at port, a truck breaks down mid-route, or a customer changes an order after the truck has already left the depot, and rebuilding that plan under time pressure is still done by dispatchers and warehouse leads, not by the optimization engine itself.
- Overview: Separate the planning layer of logistics, which is heavily software-driven, from the execution layer, which absorbs the disruptions that planning cannot predict in advance. Routing algorithms, warehouse slotting, demand forecasting, and freight-rate analysis are increasingly automated and improving each quarter as more shipment data feeds the models. Live dispatch changes, warehouse exception handling when a pallet is damaged or mislabeled, last-mile delivery problems, and carrier negotiations during a capacity crunch still depend on people who can adapt the plan in real time as conditions shift underneath it.
- Automation: AI moves first into route optimization engines, automated warehouse picking and sortation robotics, demand-forecasting models that set inventory levels across a distribution network, dynamic freight-rate pricing, and predictive delivery windows shared automatically with customers well before arrival. It stalls when a shipment is delayed at a port or border crossing, when a warehouse robot jams or a pallet arrives damaged and needs a human decision on the spot, when a driver hits a road closure the routing engine never knew about, and when a customer escalation requires renegotiating a delivery promise on the fly under real time pressure and conflicting priorities.
- Human: Durable roles include dispatchers who reroute drivers around real-time disruptions as they unfold, warehouse supervisors who resolve exceptions that robots and scanners cannot handle on their own, freight brokers who negotiate capacity during regional shortages or seasonal surges, and last-mile drivers and delivery staff who handle the unpredictable conditions of an actual address, an actual customer, and an actual loading dock. Customs and compliance specialists who untangle a held shipment also carry outsized value. These roles absorb the gap between the optimized plan and what actually happens each day.
- Reading: Score this industry by asking whether a role is mostly generating the optimized plan or mostly executing and repairing that plan when reality disrupts it without warning. Planning analysts and back-office logistics roles face faster automation pressure as forecasting tools mature and scale across more networks. Dispatch, warehouse exception-handling, and last-mile delivery roles keep more weight in the score because disruption in this industry is constant, not occasional or rare.

### Manufacturing

- URL: https://ai-job-risk.net/industries/manufacturing
- Lead: Manufacturing looks highly automatable because so much of the floor is already measured: sensors track vibration, temperature, and output rates, and computer-vision systems catch a visible defect on a line faster and more consistently than an inspector scanning the same part all shift. Predictive-maintenance systems already flag bearing wear or motor strain before a human would notice. But a factory does not run itself when something goes wrong. Deciding why a defect appeared and how to fix a process that no longer matches its specification still depends on people who understand the equipment, not just the data feed.
- Overview: Separate manufacturing work that is stable and measurable from work that responds to the unexpected. Visual inspection against a known defect pattern, logging work records, flagging predictive-maintenance alerts, and comparing a process against its standard parameters are stable: sensors and vision systems already do much of this continuously. Deciding to stop a line, diagnosing why a new kind of defect is appearing, and redesigning a process step that no longer fits conditions are response work: they require a working model of the machine and product that a monitoring system doesn't have. The same floor contains both kinds of work in different proportions depending on the role.
- Automation: AI and automation move first into visual defect inspection on the line, predictive-maintenance alerts from vibration and thermal sensors, automated work-record and production logging, statistical process comparison against specification, and scheduling support that sequences jobs across machines. Robotic arms already handle repetitive assembly and material handling on stable, high-volume lines. It stalls on stoppage decisions when a defect pattern doesn't match anything the system has seen before, root-cause diagnosis that requires tracing a problem back through several process steps, and floor-level process improvement that comes from an experienced operator noticing something a sensor isn't tuned to detect.
- Human: What stays durably human is diagnosing the unfamiliar and deciding how to act on it. Maintenance technicians who trace an intermittent fault back to its actual cause, line supervisors who decide whether a deviation is safe to run through or needs a stoppage, and process engineers who redesign a step after noticing a recurring near-miss carry judgment that sensors don't replicate. Experienced operators who can tell something is off before any gauge shows it, and quality engineers who investigate why a defect is appearing rather than just flagging that it did, keep work that automation supports but doesn't take over.
- Reading: Read manufacturing scores by separating routine monitoring from response and diagnosis. A quality-control inspector doing visual checks against a fixed standard scores higher risk because vision systems already do this reliably. A maintenance technician who diagnoses root causes or a line supervisor who makes stoppage calls scores lower because the job is responding to the unexpected, not repeating a check. A single production floor can show a wide range of scores across roles that all sound like manufacturing work from outside.

### Marketing

- URL: https://ai-job-risk.net/industries/marketing
- Lead: Marketing runs on exactly the kind of output large language models are good at: ad copy, email sequences, social captions, landing-page variants, and reporting decks. Tools that draft, resize, and A/B-test creative at scale have already changed how account teams and in-house marketers spend their day, and campaign reporting that once took an analyst hours now takes minutes. But a campaign is not just assembled words and charts. Someone still has to decide what the brand should say, which audience insight is worth acting on, and whether a message will actually land with a skeptical buyer rather than just look plausible in a slide.
- Overview: Read marketing risk by separating tasks that are mostly execution from tasks that are mostly judgment. Drafting ad variants, building keyword lists, tagging audience segments, and generating standard performance reports are execution: AI already does large parts of this, and speed gains are real and measurable in campaign turnaround times. Brand positioning, choosing which market signal matters, and deciding how a company should sound when something goes wrong are judgment calls: they require context AI does not have and consequences a model cannot be held to. The same job title can sit on either side depending on how much of the role is actually spent on each.
- Automation: AI moves first into copy drafting, subject-line and headline testing, image and video variant generation for ad sets, campaign reporting and dashboarding, audience segmentation, bid and budget optimization across channels, and first-pass keyword and competitor research pulled from search and social data. Programmatic ad platforms already handle much of the moment-to-moment bidding and targeting that media buyers used to adjust manually. It stalls on setting brand voice and guardrails, choosing a positioning strategy against competitors, reading which cultural or market moment is worth responding to, negotiating with agencies or platforms, and taking responsibility when a campaign misreads its audience or damages the brand.
- Human: What stays durably human is brand strategy, positioning, and the judgment calls that follow a campaign after it launches. Brand managers and strategists who decide what a company stands for, client leads who read a room and adjust a pitch mid-meeting, and creative directors who reject an on-brief but off-tone draft are hard to replace with a model. So are the people who interpret ambiguous market research, negotiate media contracts, and decide how to respond when a campaign draws public backlash. Roles built around audience insight, brand judgment, and cross-functional persuasion keep their value even as production work shrinks.
- Reading: Read marketing scores by asking how much of the role is production versus positioning. A performance-marketing or content-production role scores higher risk because output volume is the job. A brand strategist, senior account director, or creative director scores lower because the job is judgment about audience and message, not word count or ad variants produced per week. The same title can carry different risk depending on whether the person mostly executes campaigns or decides what campaigns should say.

### Media

- URL: https://ai-job-risk.net/industries/media
- Lead: Media work runs on words, footage, and speed, which is precisely what generative tools now do well: transcribing interviews, drafting story summaries, generating headline options, and tagging hours of raw footage in minutes instead of hours. Newsrooms and production houses that once needed a full shift to turn around a segment can now get a rough cut or a first draft almost immediately. But publishing something is not the same as knowing it deserves to be published. Deciding what is true, what is newsworthy, and whose account to trust still depends on a person who can be named and held accountable for being wrong.
- Overview: Separate media tasks that are mostly mechanical from tasks that carry editorial weight. Transcription, rough-cut assembly, metadata tagging, headline variant generation, and first-draft summarization are mechanical: AI already compresses these from hours to minutes. Deciding which story to pursue, verifying a source, framing a sensitive topic responsibly, and standing behind a claim in print or on air are editorial: they require judgment about credibility and consequence that cannot be outsourced to a model trained on pattern-matching rather than accountability. A newsroom or studio can automate the first kind of work heavily while the second kind barely moves.
- Automation: AI moves first into transcription and captioning, first-draft article and script generation from source material, footage and asset tagging, headline and thumbnail variant testing, translation and localization passes, and templated formats like market recaps or sports summaries. Archive search across large footage or article libraries, once a researcher's task, is now largely automated. It stalls on sourcing a story that no one has reported yet, building trust with a confidential source, verifying a claim against conflicting evidence, and making the editorial call on how to frame something politically or emotionally sensitive without a template to follow.
- Human: What stays durably human is sourcing, verification, and editorial judgment under pressure. Investigative reporters who cultivate sources over years, editors who decide what a front page or homepage says about an organization's values, and producers who make split-second calls about what to air during a live or breaking event are not replaceable by a drafting tool. Fact-checkers and standards editors who catch a plausible-sounding but false claim, and correspondents who build the trust needed to get someone to talk on the record, carry a kind of credibility that has to be earned by a person, not generated by a model.
- Reading: Read media scores by separating content production from editorial responsibility. A role built around transcription, formatting, or templated writeups scores higher risk because the output is repeatable and verifiable against a source. A role built around sourcing, verification, or editorial judgment scores lower because the work is deciding what to trust and what to publish, not producing text faster. Two people with the same job title, one doing rewrites and one doing original reporting, can carry very different scores.

### Operations

- URL: https://ai-job-risk.net/industries/operations
- Lead: Operations is the industry most exposed to automation on paper, because so much of the daily work is data entry, reconciliation, ticket routing, and status reporting that follows a fixed, documented procedure. Much of that volume has already moved to bots and workflow software. But operations exists precisely because processes break in ways documentation never anticipated: a shipment doesn't arrive, a system integration silently fails, two departments disagree about who owns a task. Noticing something has gone off-script and getting it back on track is what keeps operations roles from collapsing into pure automation.
- Overview: Split operations work into steady-state processing and exception handling, because the two behave completely differently under automation pressure. Data entry, invoice matching, status updates, routine ticket triage, and standard reporting follow repeatable rules and are the fastest-moving part of the job as volume shifts to software. Coordinating across teams when a process breaks down, prioritizing competing urgent requests with no clear precedent, and tracking down the root cause of a recurring problem require judgment about a specific, non-repeating situation, and that is the part of operations work that keeps its value even as routine volume shrinks steadily.
- Automation: AI and robotic process automation move first through invoice and purchase-order matching, data entry between disconnected systems, first-line ticket triage and routing, scheduling, and standard status reports that used to require a person to compile manually. Workflow tools already auto-escalate simple exceptions based on predefined rules and thresholds. It stalls when the exception doesn't match a known pattern: a vendor dispute that genuinely needs a phone call to resolve, a system outage where a human has to decide which of five fires to fight first, or a cross-team handoff where the documented process no longer matches what actually happens on the ground.
- Human: The durable roles in operations are the ones that absorb the friction automation cannot: coordinators who chase down why a process stalled somewhere between three departments, team leads who reprioritize on the fly when three urgent requests land at once, and specialists who know which undocumented workaround actually gets a stuck shipment or approval moving again. That institutional memory about how things really work, as opposed to how the flowchart says they should work, is genuinely hard to encode into any system and hard to replace with a model that has never seen the exception before.
- Reading: For operations roles, ask how much of the job is repeatable processing versus handling exceptions and coordinating people across teams. Roles built around steady, high-volume, rule-based tasks tend to score high on exposure because that volume is exactly what automation targets first. Roles that exist mainly to catch and fix what breaks, reprioritize under pressure, or coordinate across teams score lower, since exception volume tends to stay roughly constant even as routine volume automates away underneath it.

### Public Service

- URL: https://ai-job-risk.net/industries/public-service
- Lead: Public service generates enormous volumes of records, reports, and routine correspondence, and AI is well-suited to processing exactly that kind of material. Case file summarization, permit and benefits processing, dispatch support, and first drafts of routine reports are being absorbed by software in many agencies. The tension is that public service also carries decisions that must be defensible to the public: who gets emergency priority, how a case is judged, when force or intervention is warranted. Those calls carry legal and safety accountability that has to sit with an identifiable person, not a system.
- Overview: Read AI's effect in public service by separating records processing from judgment calls with consequences. Summarizing case files, searching precedent or regulation, drafting routine correspondence, and triaging incoming requests are increasingly automated. Deciding how to handle an ambiguous or high-stakes case, exercising discretion in the field, and making judgment calls that affect someone's safety or legal standing are not, because these require weighing context and consequences that a system cannot be accountable for.
- Automation: AI moves first through records and case-file summarization, benefits and permit application processing, dispatch support that helps route calls and resources, first drafts of incident and compliance reports, and search tools that surface relevant regulations or precedent. Many agencies already use automated triage for non-emergency requests. It stalls where discretion and safety meet: a caseworker judging a family's specific situation, an officer or first responder making a real-time safety call, and any decision that has to be justified afterward to a supervisor, a court, or the public as the responsible party's own judgment.
- Human: What stays durably human in public service is discretion under accountability. Caseworkers assessing an individual family's circumstances, first responders and officers making real-time safety judgments, inspectors deciding whether a specific site is genuinely compliant, and administrators exercising discretion in ambiguous cases all carry a form of accountability that has to attach to a named, responsible person rather than a system output.
- Reading: For public-service roles, weigh how much of the work is processing records and routine requests versus exercising judgment or discretion with real consequences. Roles concentrated in records processing, report drafting, or routine case intake should expect a higher score, reflecting how much of that throughput is already automatable. Roles centered on field judgment, safety decisions, or case-specific discretion should read a lower score as reflecting real accountability that has to remain human.

### Real Estate

- URL: https://ai-job-risk.net/industries/real-estate
- Lead: Real estate generates enormous amounts of structured data: listings, comparable sales, price histories, and inspection reports, all of which AI tools can search, compare, and summarize quickly. Automated valuation models and AI-drafted listing copy have already changed how agents spend their prep time before a property even goes on the market. But a property transaction is also one of the largest financial decisions most people make in their lives, and it hinges on negotiation, reading a counterparty's real position, and judgment calls made by physically walking through a building, none of which a comparables report resolves on its own.
- Overview: Separate real estate work into information assembly and deal-making, since these move at very different speeds. Pulling comparable sales, running valuation models, drafting listing descriptions, and preparing standard transaction paperwork are data tasks that speed up substantially with AI support once the property records exist digitally. Negotiating price and terms between two parties who each want something different, judging a property's true condition on a walkthrough, and reassuring a nervous buyer or seller through a high-stakes decision require reading people and physical spaces, which is where the industry's durable human core actually sits.
- Automation: AI moves first through comparable-sales analysis, automated valuation models, listing-copy generation, initial buyer inquiry responses, and document preparation for standard purchase agreements and disclosure forms. Property search platforms already do more of the matching work that agents used to perform manually by phone and email, and virtual staging tools generate marketing images without a photographer visiting the property. It stalls on negotiation between parties with genuinely conflicting interests, in-person assessment of a property's actual condition beyond what photos show, and talking a client through the emotional and financial weight of the biggest purchase or sale of their life, often under real time pressure.
- Human: What stays human in real estate is negotiation and on-site judgment: agents who read what a seller will actually accept versus what the listing price says, inspectors and appraisers who catch structural or condition problems a photo never shows, and brokers who manage a deal through financing contingencies, appraisal gaps, and last-minute disputes between buyer and seller. Property managers handling tenant conflicts, maintenance emergencies, and eviction proceedings depend on the same kind of situational judgment that resists standardization into a model.
- Reading: Read a real-estate role by asking whether it is mostly listing and comparables work or mostly negotiation and in-person judgment. Roles centered on valuation reports, listing production, and transaction paperwork show higher exposure because that output is increasingly generated automatically. Roles centered on negotiating deals, assessing physical property condition, or managing client relationships through a stressful transaction score lower, because those tasks depend on presence and trust that data tools do not substitute for.

### Retail

- URL: https://ai-job-risk.net/industries/retail
- Lead: Retail has already automated much of its back office, and AI is now extending that automation to demand forecasting, pricing, and even promotional copy, which used to require a merchandiser's judgment call. Self-checkout, inventory optimization, and algorithmic reordering are established practice in most chains. The tension is between the parts of retail that are pure logistics — moving the right stock to the right shelf at the right price — and the parts that are still relational: a floor associate helping someone choose a gift, a stylist building a client relationship, a manager solving a complaint that a policy doesn't cover.
- Overview: Read AI's effect in retail by separating logistics from the floor. Demand forecasting, automated reordering, dynamic pricing, planogram generation, and drafting promotional or product copy are now largely software-driven and improving quickly. Helping an undecided customer choose between options, merchandising a store to fit local shoppers, and resolving an angry customer's problem in person are not, because they depend on reading an individual shopper's needs rather than optimizing an average one.
- Automation: AI moves first through checkout automation and self-scanning, through demand forecasting and automated replenishment that reorders stock before a human notices a shortage, through dynamic and competitive pricing engines, and through generated product descriptions and promotional copy. Warehouse and distribution-center picking is increasingly robotic. It stalls on the sales floor: helping a specific customer who doesn't know what they want, visual merchandising that responds to how real shoppers move through a real store, and de-escalating a complaint or return dispute that a return policy alone can't resolve.
- Human: What stays durably human in retail is direct customer contact and physical judgment. Sales associates helping shoppers make decisions, visual merchandisers adapting displays to how customers actually behave in a store, store managers handling escalated complaints and staffing problems, and specialists in categories that require expertise — like electronics or fine goods — all depend on reading an individual customer or physical space, not processing a transaction.
- Reading: For retail roles, weigh how much of the job is processing transactions and inventory versus advising customers face to face and adapting a physical space to them. Roles concentrated in checkout, restocking, pricing, or copywriting should expect a higher score, since forecasting and automated reordering already handle much of that work. Roles centered on customer advice, merchandising judgment, or complaint resolution should read a lower score as reflecting real insulation from current automation.

### Science

- URL: https://ai-job-risk.net/industries/science
- Lead: Science produces enormous amounts of text and data, and AI is already changing how much of that gets processed by hand. Literature review, data cleaning, and first-pass comparison of models or methods can now be done far faster than a researcher working alone. The tension is that generating a summary of prior work is not the same as knowing which question is worth asking, and running an analysis is not the same as knowing whether its result actually means what it appears to mean. Those judgment calls sit with a scientist's training and reputation, not with the tool that produced the draft.
- Overview: Read AI's effect in science by separating information processing from the scientific judgment around it. Searching and summarizing literature, cleaning and preprocessing datasets, running standard statistical or model comparisons, and drafting routine lab records move much faster with current tools. Formulating a genuinely novel question, designing an experiment that actually tests it, and interpreting an ambiguous or surprising result do not move at the same pace, because they depend on domain judgment that a tool cannot originate.
- Automation: AI moves first through literature search and summarization across large numbers of papers, through data cleaning and preprocessing pipelines, through drafting and formatting of lab notebooks and routine reports, and through rapid first-pass comparison of models, parameters, or analysis approaches. Some experimental workflows already use automated instruments for repetitive sample processing. It stalls on the core of research: deciding which question is worth pursuing, designing an experiment or study that isolates the right variable, judging whether a surprising result reflects reality or an artifact, and taking responsibility for a claim published under a researcher's name.
- Human: What stays durably human in science is originating and judging, not processing. Principal investigators and researchers framing which questions matter, experimentalists designing studies that control for the right variables, scientists interpreting ambiguous or conflicting results in context, and peer reviewers and mentors exercising judgment about what counts as a valid claim all depend on domain expertise and accountability that a model can assist but not replace.
- Reading: For science roles, weigh how much of the work is literature search, data processing, or routine analysis versus posing questions and interpreting results. Roles concentrated in literature review, data cleaning, or standard model comparisons should expect a higher score, since those tasks are already being compressed by current tools. Roles centered on original question framing, experimental design, or interpreting ambiguous findings should read a lower score as reflecting judgment that remains squarely human.

### Technology

- URL: https://ai-job-risk.net/industries/technology
- Lead: Technology is unusual among industries because it is simultaneously the source of the automation wave and one of the fields most affected by it. Code completion, test generation, and log triage are now routine parts of a developer's daily toolkit, and they have measurably sped up the mechanical parts of building software. What has not gotten easier is deciding what to build in the first place, how a system should be structured so it doesn't collapse under its own complexity two years later, and who is responsible when a production system fails at three in the morning with customers watching.
- Overview: Separate technology work into code production and system responsibility, since AI reshapes them very differently. Writing boilerplate, generating unit tests, completing routine functions, and summarizing logs for debugging are tasks that AI tools now do quickly and reasonably well, often faster than a junior engineer would. Choosing an architecture that will hold up as a product scales, deciding what tradeoffs a system should make under real constraints, and being the person who gets paged when something breaks in production require judgment about consequences that unfold over months or years, and that is where engineering seniority continues to matter most.
- Automation: AI moves first through code completion, boilerplate generation, unit and integration test writing, log and error triage, and first-draft technical documentation that engineers used to postpone indefinitely. Code review assistants already catch a meaningful share of style, correctness, and security issues before a human reviewer ever looks at a pull request. It stalls on system design decisions, debugging genuinely novel production incidents with no precedent in the runbooks, negotiating technical tradeoffs with product and business stakeholders who don't share an engineering vocabulary, and taking on-call responsibility for a system's uptime.
- Human: Durable technology roles are the ones carrying design and operational responsibility: architects who decide how services should be decomposed and where the seams should go, senior engineers who make the call during a live incident about what to roll back and what to leave alone, and staff engineers who negotiate tradeoffs between speed, cost, and reliability across competing teams. Security engineers assessing genuinely new threats and engineering managers navigating team conflict and stakeholder pressure also depend on judgment that doesn't reduce to pattern completion.
- Reading: Score a technology role by asking whether it is mostly writing code to a known spec or mostly making decisions a team and a business depend on. Roles built around routine implementation, test writing, and boilerplate work show higher exposure because AI assistants already handle much of that volume. Roles centered on system design, incident response, and technical tradeoffs under real operational stakes score lower, since AI assistance speeds up the typing without removing the responsibility for what actually gets shipped.

### Transportation

- URL: https://ai-job-risk.net/industries/transportation
- Lead: Transportation already runs on automation in many places: GPS routing, traffic-adaptive signal timing, and driver-assist systems in trucks and transit fleets all speed up daily operations across networks. Autonomous vehicle pilots are expanding in some markets as the technology matures. But moving people and freight safely still means responding to a child running into a street, an icy bridge a sensor did not flag in time, or a mechanical failure at highway speed, and that responsibility currently sits with a licensed operator whose judgment is trusted precisely because it is accountable.
- Overview: Separate the scheduling-and-routing side of transportation, which is increasingly automated, from the safety-critical operation of vehicles and vessels in unpredictable conditions on the ground, in the air, or at sea. Dispatch scheduling, traffic-flow optimization, fleet-maintenance forecasting, and fare or freight-rate systems respond well to AI support today and continue to improve as more operational data accumulates. Driving, flying, or piloting through weather, mechanical failure, traffic incidents, and passenger emergencies still requires a licensed human who bears direct personal responsibility for the final outcome.
- Automation: AI moves first into traffic-signal optimization across congested corridors, transit and freight scheduling, predictive maintenance for vehicle and rail fleets, driver-assist and collision-avoidance systems, and route planning that accounts for congestion and weather patterns well ahead of departure. It stalls on emergency maneuvers in live traffic, safe operation in severe weather or sudden mechanical failure, passenger incidents that need immediate human judgment on board, boarding disputes that de-escalate only with a person present, and the accountability chain that regulators and courts still attach to a licensed operator rather than to a system or its vendor.
- Human: Roles that stay durably human include pilots, ship captains, train engineers, and commercial drivers who remain legally and practically responsible for safe operation under abnormal conditions, along with dispatchers and air traffic controllers who make real-time safety calls during active disruptions. Transit and rail operations staff who manage a system-wide incident, such as a signal failure or a stalled train, carry similar weight. Their value lies in accountable judgment during the moments automation is least tested: failures, emergencies, and edge cases that fall well outside the training data.
- Reading: Read scores here by separating scheduling and back-office transportation roles, which automate quickly as software matures and adoption widens across carriers, from vehicle and vessel operation, which remains anchored in licensed human responsibility and regulatory oversight. The more a role is defined by responding to unpredictable conditions in motion, at speed, or with passengers aboard, the less it should be read as a straightforward automation case.

## Country Editorial Context

### Australia

- URL: https://ai-job-risk.net/countries/australia
- Lead: Australia's economy rests on large mineral and energy exports concentrated in remote regions, alongside a services economy clustered in a few coastal capital cities, with vast distances separating the two. That geography shapes AI exposure directly: office-based finance, insurance, and professional services in Sydney and Melbourne face faster automation pressure, while mining operations, agriculture, and a growing aged-care and healthcare workforce spread across the country depend on physical presence that AI cannot substitute for.
- Overview: Australia is best read by separating city-based knowledge and service work from resource extraction and care work spread across a huge, sparsely populated landmass. Banking, insurance, and professional services concentrated in state capitals sit in the exposed layer, since this work is standardized, digitized, and unconstrained by distance. Mining operations, agriculture, and the fast-growing aged-care and disability-support sectors sit in the durable layer, where site-based safety obligations, seasonal and environmental variability, and direct personal care keep human labor central regardless of how advanced supporting software becomes.
- Sector: Employment concentrates in financial and professional services in the major capitals, a mining and resources sector that drives exports but employs a smaller, geographically remote workforce, and a large and expanding health and aged-care sector responding to an ageing population. AI pressure is strongest in mortgage processing, insurance claims, superannuation administration, and back-office banking. It moves more slowly through fly-in fly-out mining operations, farming tied to land and weather, and direct aged-care and disability support work, where an ageing population is increasing demand for hands-on human care faster than automation can substitute for it.
- Resilience: The most durable roles combine site-based physical work with rising demand from demographic change: mining technicians and heavy-equipment operators, agricultural workers tied to seasonal and regional conditions, and aged-care and disability-support workers whose numbers need to grow simply because Australia's population is getting older. The sheer distance separating major cities from resource regions also means centrally designed automation is harder to roll out uniformly across the country, and remote and fly-in fly-out work schedules keep a premium on people who can be physically present on site.
- Limits: A national score blends a handful of dense coastal service economies with vast inland mining and agricultural regions, plus a care sector expanding for demographic reasons unrelated to AI. It cannot show how much of the measured exposure sits in Sydney and Melbourne offices versus remote mine sites and farms hundreds of kilometers away. Read the score alongside geographic concentration and the demand pressure coming from an ageing population, not as a single uniform trend across such a large country.

### Brazil

- URL: https://ai-job-risk.net/countries/brazil
- Lead: Brazil's labor market is defined by a large agribusiness export sector, a sprawling services economy, and one of the largest informal workforces of any major economy, alongside a banking and fintech sector that has industrialized much faster than the rest of the country. That split matters for AI: digital banking, call centers, and back-office finance are already deeply exposed to automation, while the vast informal and semi-formal service economy, agricultural fieldwork, and small-scale retail run on cash, personal trust, and physical presence that software cannot reach.
- Overview: Reading Brazil well means separating the formal, digitized economy from the much larger informal one. Banks, insurers, telecoms, and large retailers have invested heavily in automated customer service, credit scoring, and back-office processing, so clerical and support roles inside those firms face real AI pressure. Outside that formal core, informal commerce, domestic services, construction labor, and small agricultural operations employ a huge share of workers in ways that are barely touched by digital systems, so a single national score has to be read against how much of the workforce sits in each layer.
- Sector: Brazil concentrates employment in agribusiness (soy, beef, sugar, coffee), retail and wholesale trade, a large public sector, construction, and a banking and fintech industry that is unusually advanced for the region and widely used even by lower-income households. AI pressure is heaviest in banking operations, insurance underwriting, call-center and customer-support work, and standardized administrative processing in large firms and government agencies. It moves far more slowly through agricultural fieldwork, logistics on poor or informal road networks, construction trades, and the informal retail and service stalls that make up much of urban commercial life across the country.
- Resilience: What holds up in Brazil is work tied to physical terrain, informal-economy trust, and direct human negotiation. Agricultural labor, equipment operation on farms, construction trades, and last-mile delivery in dense or informal urban areas depend on physical adaptability and local knowledge that software cannot supply. Much of the services economy also runs on personal relationships and cash transactions built on trust between neighbors and repeat customers, a structure that resists standardization regardless of how capable AI tools become inside the formal sector.
- Limits: A single national score for Brazil compresses a highly formalized banking and corporate sector into the same number as a very large informal economy that operates by different rules entirely. Regional gaps between wealthier southern industrial states and less-formalized northern and interior regions add another layer the average cannot show. Read the score together with how much of a given local labor market sits inside registered, digitized firms versus informal, cash-based work.

### Canada

- URL: https://ai-job-risk.net/countries/canada
- Lead: Canada pairs an advanced, urban knowledge economy in finance, technology, and professional services with a resource economy built on energy, mining, and forestry that operates across vast, sparsely populated territory. That geography matters as much as sector mix: AI adoption moves fastest in the office towers of Toronto, Montreal, and Calgary, while extraction, transportation, and the bilingual federal and provincial public service depend on physical operations and language-specific administrative work that resist quick automation.
- Overview: Canada is easiest to read by separating concentrated urban knowledge work from resource extraction and the public administration spread across the country. Banking, insurance, IT services, and corporate administrative work sit in the exposed layer, concentrated in a handful of major metro areas with deep financial and tech sectors. Mining, oil and gas field operations, forestry, and the large bilingual public service sit in the more durable layer, where remote site work, safety-critical operations, and French-English service obligations under federal law keep substantial human staffing in place.
- Sector: Employment concentrates in financial services and technology in central Canada, energy and mining in Alberta and parts of the west, forestry and fisheries in coastal and northern regions, and a large public sector shaped by federal bilingualism requirements. AI pressure is sharpest in bank back offices, insurance underwriting, IT support, and administrative processing in major cities. It spreads more slowly through oil sands and mine site operations, forestry and fisheries work tied to remote regions, and public-facing government services that must be delivered in both English and French, which adds translation and compliance layers that slow any purely automated rollout.
- Resilience: What stays durable is work anchored to physical resource operations and to Canada's bilingual public administration: mine and rig technicians, foresters, fisheries workers, and civil servants delivering federally mandated services in two languages. The country's reliance on a resource sector operating far from major population centers also means site supervision, equipment maintenance, and safety judgment retain outsized importance relative to economies built more heavily around office-based services, and translation obligations add a layer of work that has no equivalent in unilingual economies.
- Limits: A single Canadian score blends a small number of dense, AI-exposed metro economies with vast resource and rural regions where the labor market looks entirely different. It also cannot capture how bilingual service obligations slow automation in public administration specifically in Canada, a factor with no equivalent in most peer economies. Read the number together with the urban-resource divide and provincial variation, rather than as one uniform national trend.

### China

- URL: https://ai-job-risk.net/countries/china
- Lead: China combines the world's largest manufacturing base with a fast-moving domestic tech sector and platform economy, so AI capability spreads through factories, logistics networks, and e-commerce operations unusually quickly. The core tension is between a huge, mobile workforce still doing physical assembly, delivery, and construction work, and an equally large layer of standardized office, customer-service, and data-entry roles inside factories, platforms, and state institutions that domestic AI tools can absorb rapidly given the scale of adoption already underway.
- Overview: China is easiest to read by separating physical, distributed labor, factory-floor assembly, warehouse and delivery logistics, construction, from the office and platform-support layer sitting on top of it. Domestic AI adoption is aggressive across e-commerce customer service, content moderation, financial back offices, and administrative functions inside both state-owned enterprises and private tech firms. That pressure is real but concentrated; it moves much more slowly through last-mile delivery, skilled factory-line work, and construction, where China's enormous migrant and informal workforce still performs tasks that are cheap, flexible, and physically grounded.
- Sector: China's employment base spans electronics and heavy manufacturing, a vast logistics and e-commerce delivery network, construction, and a rapidly digitizing services and finance sector concentrated in major cities. AI pressure is heaviest in platform-company back offices, customer support, financial processing, translation, and the standardized reporting demanded by both corporate and government bureaucracy, where domestic large language models are being deployed at scale. It is far lighter across delivery riders, factory-line operators, construction crews, and the broader informal and migrant-labor economy, where low labor costs, physical requirements, and localized coordination keep automation economically or practically unattractive for now.
- Resilience: The most durable work in China sits in physical logistics, skilled manufacturing supervision, and construction, sectors that employ enormous numbers of migrant workers and depend on flexible, low-cost physical labor that remains cheaper and more adaptable than automation in many contexts. Roles that manage exceptions on fast-scaling platforms, quality inspection on production lines, and coordination between suppliers, factories, and delivery networks also hold value because they require judgment about fast-changing, real-world conditions that centralized systems still struggle to fully standardize.
- Limits: A national figure cannot reconcile China's uneven development: coastal megacities with mature digital infrastructure sit alongside inland regions where manual and informal labor still dominate. State industrial policy and rapid domestic AI rollout can also shift adoption speed faster than market signals alone would suggest. Read the score together with the rural-urban divide, the size of the migrant labor force, and the gap between platform-company white-collar work and physical logistics or manufacturing labor.

### France

- URL: https://ai-job-risk.net/countries/france
- Lead: France combines a large state sector and strong labor protections, including strict dismissal rules and influential works councils, with a genuine industrial base in aerospace, automotive, luxury goods, and agriculture. That legal and institutional structure slows how quickly AI-driven efficiency gains convert into actual job losses, even in roles where the underlying work, such as administrative processing or standardized analysis, is highly exposed to automation in principle, because employers cannot simply cut staff the way they might in a more lightly regulated labor market.
- Overview: France is best read by separating what AI can technically do from how quickly French labor law and institutions allow that capability to change staffing. Administrative processing, standardized financial analysis, and routine drafting sit in the exposed layer on a purely technical basis. Public-sector employment, unionized industrial roles, and regulated professions sit in the durable layer for institutional as well as technical reasons: strong dismissal protections, mandatory employee consultation through works councils, and civil-service status all raise the practical cost and legal complexity of converting automation into job cuts.
- Sector: Employment concentrates in a large public administration and public services sector, aerospace and automotive manufacturing, luxury goods and agri-food production, and tourism. AI pressure is strongest in banking back offices, insurance processing, and standardized corporate administrative work, where technical exposure is high. It moves more slowly through the state sector, where civil-servant status limits at-will restructuring, through unionized manufacturing plants in aerospace and automotive where works councils must be consulted on major changes, and through agriculture and tourism, which depend on physical presence and seasonal, regional variation.
- Resilience: What holds up best is work protected by institutional structure as much as by task content: civil servants, unionized industrial roles in aerospace and automotive manufacturing, and regulated professions such as notaires and medical practitioners whose functions are defined by law. France's strong collective-bargaining tradition and works-council consultation requirements mean that even when automation is technically ready, employers face real procedural constraints before headcount actually changes, which slows displacement even in administrative roles that are otherwise highly exposed.
- Limits: A national score captures technical exposure well but understates how much French labor law and collective bargaining slow the conversion of exposure into actual job change, especially in the public sector and unionized industries. It also blends Paris-centered service and administrative work with regional manufacturing and agricultural employment that faces a different mix of pressures. Read the number alongside sector mix and the strength of employment protection, not as a direct forecast of job losses.

### Germany

- URL: https://ai-job-risk.net/countries/germany
- Lead: Germany's labor market is shaped by a large industrial base, a dense network of mid-sized manufacturers — the Mittelstand — and a heavily documented, process-driven administrative culture. That combination makes parts of the economy very exposed to AI support, especially clerical, back-office, and standardized analytical work, while skilled trades, engineering judgment, and precision manufacturing hold up better because they depend on physical work and hard-won expertise that is not easily reduced to text.
- Overview: Germany is easiest to read by separating the highly documented office and administrative layer from the engineering, skilled-trade, and production work that anchors the economy. AI can accelerate a large share of clerical processing, translation, reporting, and first-pass analysis, and Germany has plenty of that work. But the country's competitive strength sits in areas — mechanical and electrical engineering, industrial maintenance, quality assurance — where responsibility, on-site judgment, and deep domain skill still decide outcomes, so a single national number hides very different realities.
- Sector: Germany concentrates employment in automotive and machinery manufacturing, engineering services, chemicals, finance and insurance, and a large public administration. AI pressure rises fastest in the office-heavy layers: administrative processing, bookkeeping, customer correspondence, standardized reporting, and translation, all of which are text- and rule-based. It moves more slowly through the industrial core, where roles such as machine setters, maintenance technicians, and quality engineers depend on physical presence, safety accountability, and judgment about equipment and tolerances that generative tools cannot take over.
- Resilience: The most durable work in Germany tends to combine technical skill with responsibility a person has to carry: engineers who sign off on designs, technicians who keep production lines running, tradespeople in construction and installation, and skilled workers whose vocational training through the dual system builds expertise that is hard to compress. Roles that pair domain knowledge with on-site decision-making and accountability keep more of their value than roles built mainly on moving documents.
- Limits: A national score cannot capture the gap between Germany's exposed administrative layer and its resilient industrial and skilled-trade base. Strong labor protections and works councils also shape how quickly change actually reaches jobs, separately from how technically automatable a task is. Read the number alongside how much of a given role is standardized office work versus physical, technical, or accountable work.

### India

- URL: https://ai-job-risk.net/countries/india
- Lead: India's labor market carries a uniquely exposed layer: a massive IT-services, call-center, and business-process-outsourcing industry built on exactly the kind of standardized coding, document processing, and customer-support work that large language models now perform directly. That exposure sits alongside a vast informal economy, agriculture, and domestic services sector that operates almost entirely outside the reach of AI tools, making India's labor market one of the most internally divided in the world.
- Overview: India is best read by separating its globally connected IT-services and BPO export sector, where AI substitution is direct and already underway, from the much larger domestic economy of informal work, agriculture, retail, and local services that AI barely touches. Junior software testing, routine coding, tier-one call-center support, and back-office outsourcing work done for Western clients face immediate competitive pressure from generative AI tools that can now perform many of the same tasks. That pressure has little bearing on the hundreds of millions of workers in agriculture, construction, small retail, and informal services who make up most of India's actual employment.
- Sector: India's economy combines a large, export-oriented IT-services and business-process-outsourcing sector concentrated in cities like Bengaluru, Hyderabad, and Pune, alongside agriculture and a vast informal economy that still employs the majority of the workforce. AI pressure concentrates heavily on entry-level and mid-tier IT work: routine software testing, basic coding, tier-one technical support, and back-office BPO processing for international clients are the most directly exposed roles anywhere in India's economy. That pressure barely reaches agriculture, construction, domestic retail, and informal services, which remain labor-intensive, cash-based, and largely untouched by enterprise AI adoption.
- Resilience: What remains durable in India spans two very different registers. In the formal IT sector, senior engineers, architects, and client-relationship managers who translate ambiguous business needs into technical solutions retain value that junior, task-execution roles do not. Across the much larger informal economy, agricultural labor, construction, small-scale retail, and household and personal services remain resilient simply because they are physical, cash-based, and outside the reach of software tools entirely, regardless of how advanced AI becomes.
- Limits: A single country score for India is unusually misleading because it averages a small, globally exposed IT-and-BPO sector against an enormous informal economy that operates by entirely different rules. The same number cannot describe both a Bengaluru software campus and a rural agricultural household. Read the score together with the sharp divide between formal, export-facing digital work and the vast informal and agricultural labor force that makes up most of India's actual employment.

### Italy

- URL: https://ai-job-risk.net/countries/italy
- Lead: Italy's economy runs on a dense fabric of small and mid-sized family firms organized into regional manufacturing districts, alongside a tourism sector built on food, fashion, and cultural heritage, and a public administration that is large and procedural. That mix creates uneven AI exposure: administrative paperwork, standardized accounting, and routine correspondence are readily assisted by AI, while artisanal manufacturing, design-led fashion work, and hospitality built on personal service resist automation because they depend on craft skill, taste, and trust built over years.
- Overview: Reading Italy well means separating the office layer from the workshop floor. A large share of employment sits inside small, owner-operated firms in textiles, leather goods, furniture, and food production, where decisions are informal and tacit rather than documented in systems AI can absorb. Alongside this, banks, insurers, and public offices generate the kind of structured, repetitive paperwork that AI tools handle comfortably. The result is a labor market where automation pressure concentrates in back-office and administrative layers while production and service work tied to regional craft traditions changes far more slowly.
- Sector: Italy's employment concentrates in manufacturing districts specializing in fashion, leather, furniture, and machinery, alongside tourism, hospitality, food service, and a sizeable public sector. AI pressure moves fastest through banking back offices, insurance claims processing, public administration paperwork, and standardized bookkeeping in the many small firms that outsource such tasks to shared services. It moves far more slowly through the artisanal core, where roles like pattern makers, master tailors, furniture craftsmen, and food producers rely on manual skill and design judgment, and through tourism and hospitality roles where guests expect personal attention, local knowledge, and hands-on service that no software replicates.
- Resilience: What holds up in Italy is work anchored in craft, design sense, and personal relationships built inside a family or district-based business model. Artisans, designers, and small-firm owners who combine technical skill with aesthetic judgment remain difficult to replace, as do hospitality and food-service roles where personal familiarity defines the customer experience. In a country where thousands of small firms operate with informal, relationship-based management rather than standardized corporate processes, the kind of large-scale, rules-based automation that works well in bigger organizations has less to grab onto.
- Limits: A single national score cannot capture the gap between Italy's industrial north, with its dense manufacturing districts, and its more tourism- and agriculture-dependent south, or between large firms and the vast population of family-run micro-enterprises that operate outside standardized corporate systems. Read this score as a broad signal, not a claim that clerical work in a Milan bank and artisanal work in a Tuscan workshop face the same pressure.

### Japan

- URL: https://ai-job-risk.net/countries/japan
- Lead: Japan's labor market rests on a large manufacturing and automotive base, a vast layer of clerical and administrative work, and a business culture built around detailed process discipline, seniority, and consensus. That combination makes back-office paperwork, standardized reporting, and routine customer correspondence highly exposed to AI support, while an aging, shrinking workforce keeps demand for people high in caregiving, skilled trades, and on-site manufacturing roles that depend on physical presence and accumulated craft knowledge rather than text-based reasoning.
- Overview: Japan is best read by separating the country's dense administrative and clerical layer, still often built around paper forms, stamps, and manual approval chains, from the manufacturing floor and service front lines where physical presence and accountability still dominate. AI spreads quickly through document processing, translation, scheduling, and internal reporting inside large corporations and government offices. It moves far more slowly through automotive and electronics production, retail floor operations, and the elder-care sector, where labor shortages driven by demographic decline mean many roles are becoming harder to fill, not easier to automate away.
- Sector: Japan concentrates employment in automotive and electronics manufacturing, a large keiretsu-style supplier network, retail and convenience-store services, and an outsized public and corporate administrative sector known for meticulous paperwork. AI pressure concentrates in back-office functions: general affairs, accounting, HR administration, translation, and the standardized reporting that fills much of Japanese white-collar work. It spreads more slowly across the shop floor and service counter, where line workers, maintenance technicians, retail staff, and caregivers depend on physical tasks, customer-facing courtesy norms, and the kind of tacit, on-the-job knowledge that Japanese firms have historically prized over formal credentials.
- Resilience: What holds up best in Japan is work anchored in physical craft, quality assurance, and caregiving, areas already strained by a shrinking working-age population. Manufacturing roles built on monozukuri-style precision and defect-catching, skilled trades, and the elder-care and nursing workforce serving one of the world's oldest populations remain durable because they require hands-on presence and trust that cannot be replicated by software. Middle-management coordination, still central to consensus-based decision-making, also resists full automation even as the paperwork underneath it shrinks.
- Limits: A single national score cannot capture the split between Japan's shrinking pool of routine clerical roles, its labor-short caregiving and skilled-trades sectors, and its export-facing manufacturing core. Lifetime employment norms and gradual internal redeployment also mean measured exposure translates into job loss more slowly here than in markets with weaker employment protections. Read the score alongside sector mix and the demographic pressure that is separately driving demand for workers regardless of automation.

### Mexico

- URL: https://ai-job-risk.net/countries/mexico
- Lead: Mexico's labor market centers on export manufacturing, particularly automotive and electronics assembly along the US border, alongside a large services sector and a wave of nearshoring investment as companies relocate supply chains closer to the United States. That combination creates a split exposure: administrative, planning, and logistics-coordination work tied to manufacturing is increasingly assisted by AI, while the assembly-line and skilled-trade work that nearshoring is expanding remains physically grounded and much harder to automate away.
- Overview: Mexico is best read as two economies moving at different speeds. The maquiladora and automotive-manufacturing belt along the northern border is deeply integrated into North American supply chains, and the office layer around it, purchasing, quality documentation, logistics scheduling, customer communication, is increasingly exposed to AI-assisted tools. The plant floor itself, along with the country's large informal and small-business services sector, changes much more slowly, since it depends on manual dexterity, physical inspection, and face-to-face commerce that remain largely outside what generative tools can take over.
- Sector: Mexico concentrates employment in automotive and electronics manufacturing, maquiladora assembly, logistics tied to cross-border trade, retail and wholesale commerce, tourism, and a large informal services sector. AI pressure concentrates in supply-chain planning, procurement documentation, customer support, and standardized back-office work at manufacturers and their corporate suppliers, especially those integrated with US and Canadian partners. It moves more slowly on the factory floor, where assemblers, machine operators, and quality inspectors depend on manual precision and physical judgment, and in the informal commerce and tourism-service jobs that make up much of urban employment.
- Resilience: The roles that hold up best in Mexico are grounded in physical assembly, hands-on maintenance, and direct customer service. Line workers, equipment technicians, and quality inspectors in the manufacturing belt remain essential because their work involves physical manipulation and situational judgment on the shop floor, not paperwork. Tourism, hospitality, and the informal retail and food-service economy also depend on face-to-face interaction and local trust that AI tools do not replace, and nearshoring investment is, if anything, adding more of this physical-production work rather than reducing it.
- Limits: A national score blends a highly export-oriented manufacturing corridor near the US border with services and informal-economy work spread across the rest of the country, two labor markets with very different exposure profiles. Nearshoring is actively reshaping demand in ways a static score cannot fully capture, since it is adding physical production jobs even as it professionalizes the administrative layer around them. Read the score alongside a region's specific tie to manufacturing versus domestic services.

### Netherlands

- URL: https://ai-job-risk.net/countries/netherlands
- Lead: The Netherlands runs one of Europe's most digitalized economies, built around a globally significant logistics and port infrastructure centered on Rotterdam, an intensive and highly mechanized agri-food sector, and a large finance and business-services cluster in Amsterdam. High digital maturity means AI tools are adopted quickly here, but it also means much of the routine, systemizable work has already been streamlined over decades, leaving remaining roles concentrated in physical logistics, specialized agricultural production, and judgment-heavy financial and advisory work.
- Overview: The Netherlands is best understood as an economy where digital efficiency is already the norm, so the marginal effect of AI falls more on incremental gains in already-optimized office work than on wholesale disruption. Its port and logistics operations combine advanced automation with large numbers of physical roles in freight handling and transport coordination that still require human oversight. Its agri-food sector depends on specialized technical knowledge in greenhouse and dairy operations. Reading the country well means recognizing that high digital readiness cuts both ways: rapid AI adoption, but also a labor market that has already absorbed much of the easy efficiency gains.
- Sector: Dutch employment concentrates in logistics and transport around the port of Rotterdam and Schiphol air cargo, high-intensity horticulture and dairy farming, finance and business services in Amsterdam, and a substantial public and healthcare sector. AI pressure is most visible in financial back offices, insurance administration, and standardized consulting deliverables, built on structured data the country's mature digital infrastructure makes easy to process. It is slower in port and warehouse operations, where physical freight handling still needs people on-site, and in agriculture, where climate-controlled greenhouse and livestock operations depend on specialized technical judgment generic AI tools do not replicate.
- Resilience: What remains durable in the Netherlands is physical logistics coordination, specialized agricultural expertise, and the advisory layer of finance and business services that depends on client relationships and regulatory judgment rather than routine processing. Port and transport roles that require real-time coordination of physical goods across a dense international hub keep their value, as does the technical knowledge embedded in Dutch horticulture and dairy farming, refined over generations and tied to specific growing and breeding conditions.
- Limits: A national score cannot fully capture how much of the Netherlands' apparent AI exposure reflects a labor market that was already highly digitalized before generative AI existed, meaning some office-efficiency gains commonly attributed to AI were realized years earlier through older software. Read this score alongside the distinction between the country's globally exposed finance and trade hubs and its more physically anchored logistics and agricultural base, which respond to new AI tools on a much slower timeline.

### Saudi Arabia

- URL: https://ai-job-risk.net/countries/saudi-arabia
- Lead: Saudi Arabia's labor market is still anchored in oil and its supporting industries, but a national transformation agenda is deliberately pushing investment into tourism, entertainment, finance, and giant construction and infrastructure projects, alongside a public sector that has traditionally employed a large share of citizens. That mix means AI adoption is arriving fastest in the administrative and financial layers tied to reform initiatives, while oil-field operations, construction, and the buildout of new cities remain heavily dependent on physical labor and specialized technical crews.
- Overview: Saudi Arabia reads most clearly as an oil-and-construction economy layered with a fast-growing, state-directed push into new sectors. Government ministries, banks, and companies tied to giga-projects have adopted AI-assisted tools for administration, reporting, and citizen services relatively quickly, reflecting deliberate national policy. At the same time, oil extraction and processing, the vast construction effort behind new cities and infrastructure, and a public sector still central to citizen employment depend on physical operations and administrative roles that are only partially touched by that same policy push.
- Sector: Saudi Arabia concentrates employment in oil and petrochemicals, construction and infrastructure development, public administration, retail, and a fast-expanding tourism, entertainment, and hospitality sector tied to national diversification goals. AI pressure concentrates in government administrative processing, banking and financial services, and corporate reporting functions, areas the national transformation strategy specifically targets for modernization. It is far lighter in oil-field operations, refining and petrochemical plant work, and the large-scale construction trades building new infrastructure, all of which depend on specialized technical skill and physical presence on site.
- Resilience: What holds up in Saudi Arabia is work tied to energy-sector operations, large-scale construction, and the physical rollout of the country's development agenda. Oil and petrochemical plant operators, maintenance technicians, and safety personnel remain essential because equipment monitoring and emergency response require people on site, not just data review. The enormous construction effort behind new cities and tourism infrastructure also depends on skilled trades and site supervision that cannot be virtualized, and hospitality roles tied to the growing tourism sector depend on direct guest service.
- Limits: A national score for Saudi Arabia has to combine a still-dominant oil and construction economy with a public sector and financial industry being deliberately modernized through top-down policy, two very different change dynamics. The pace of diversification also varies by region and by how directly a sector connects to national transformation initiatives versus traditional oil-economy employment. Read the score alongside how exposed a given sector is to policy-driven modernization versus oil, construction, and site work.

### Singapore

- URL: https://ai-job-risk.net/countries/singapore
- Lead: Singapore packs a financial hub, one of the world's busiest transshipment ports, and a compact, highly educated professional-services economy into a small city-state, so AI tools reach banking back offices, trade documentation, and corporate services with unusual speed. The core tension is between abundant, easily automated processing work in finance and logistics administration, and the country's continued reliance on regulatory precision, port and supply-chain operations, and a multinational business-hub role that still requires trusted human judgment.
- Overview: Singapore is easiest to read by separating high-volume financial and trade-document processing from the physical and regulatory work anchored around its port and regional headquarters role. As a dense financial center serving regional clients, back-office banking, compliance-document review, and administrative reporting are prime candidates for AI support. That does not extend as cleanly to port and maritime logistics, where physical cargo handling and vessel coordination stay grounded in the real world, or to the regulatory and cross-border legal work that keeps multinational firms anchored here for its reputation for reliable oversight.
- Sector: Singapore's economy concentrates in banking and wealth management, port operations and maritime logistics, business and legal services supporting regional headquarters, and a growing technology sector. AI pressure is heaviest in banking operations, compliance documentation, trade-finance processing, and corporate administrative services, where large multinational back offices already run heavily standardized workflows. It is lighter in port and terminal operations, where physical container handling and vessel scheduling require on-the-ground coordination, and in the regulatory, tax, and legal advisory work that draws multinational firms to base regional operations in Singapore for its institutional predictability.
- Resilience: What stays durable in Singapore is work built on regulatory trust and physical trade infrastructure. Compliance officers, cross-border legal and tax advisors, and relationship-focused private bankers remain valuable because Singapore's appeal as a business hub rests on institutional credibility that requires accountable human sign-off. Port operations, maritime logistics coordination, and supply-chain management roles also hold up because moving physical goods through one of the world's busiest ports still depends on real-time, on-site coordination.
- Limits: A single national score compresses a genuinely dual economy: a small, highly automatable financial and administrative core, and a physical logistics and port sector that is comparatively insulated. Singapore's small size and high skill base also mean displaced back-office workers can be redeployed within the same tight labor market faster than in larger, more fragmented economies. Read the score together with the balance between financial-sector processing work and the port-and-trade infrastructure that anchors the physical economy.

### South Korea

- URL: https://ai-job-risk.net/countries/south-korea
- Lead: South Korea's economy is built around a small number of dominant conglomerates in semiconductors, electronics, and automotive manufacturing, layered onto one of the most digitized and densely connected societies in the world. That makes structured office work, especially administrative and support roles inside these chaebol, highly exposed to AI, while the country's intensely credential-driven education culture and its precision manufacturing base, where fabrication yield and hardware reliability carry enormous stakes, keep specific technical and quality roles firmly in human hands.
- Overview: South Korea is best read by separating semiconductor and electronics manufacturing, where physical process control and yield management dominate, from the administrative and reporting layers inside the chaebol and their supplier networks. High broadband penetration and near-universal digital literacy mean AI tools spread through office work, translation, and customer service unusually fast. That speed does not carry over cleanly into fabrication plants and hardware engineering, where defect rates depend on specialized technicians, nor into the tutoring-heavy private education sector, where parental trust and exam outcomes still favor human instructors.
- Sector: Employment concentrates in semiconductor fabrication, electronics and automotive manufacturing, shipbuilding, and a services sector anchored by finance, telecom, and a large private education and tutoring industry (hagwon). AI pressure is strongest in administrative and clerical work inside large conglomerates, standardized financial processing, translation, and customer support. It is weaker in semiconductor fab operations, where process engineers and equipment technicians manage yield and contamination control, in shipbuilding and heavy industry requiring physical precision, and in the private tutoring sector, where results-driven parents still place a premium on human instruction ahead of national exams.
- Resilience: What remains durable in South Korea is work tied to hardware reliability, physical process control, and reputation-sensitive service. Semiconductor process engineers, quality technicians in electronics and automotive plants, and skilled tradespeople in shipbuilding hold their value because equipment failures and yield problems carry outsized financial consequences that still require expert human diagnosis. Exam-focused private tutors and counselors also remain resilient given how central university entrance outcomes are to family life and social standing.
- Limits: A single score glosses over the gap between Seoul's highly digitized corporate and service sectors and the industrial cities built around shipyards, fabs, and auto plants, where physical process work dominates. It also cannot capture how conglomerate hiring practices and strong internal labor protections slow the conversion of automation exposure into actual job loss. Read the score together with the split between office-based conglomerate work and hardware-dependent manufacturing roles.

### Spain

- URL: https://ai-job-risk.net/countries/spain
- Lead: Spain's labor market leans heavily on tourism and hospitality, a large services sector, and agriculture with strong seasonal hiring patterns, alongside growing logistics and business-services activity in its major cities. That structure means AI exposure concentrates in office and administrative work, while the country's substantial front-line hospitality, seasonal agricultural labor, and in-person service workforce remains comparatively insulated, since guest-facing and field-based work depends on physical presence and adaptability that software cannot supply.
- Overview: Spain is best read by separating its large seasonal and in-person workforce from its growing office-based services layer. Tourism alone supports a wide range of jobs in hotels, restaurants, and travel services that revolve around live human interaction, while agriculture still depends on seasonal field labor tied to harvest cycles and weather. Meanwhile, a maturing business-services and finance sector in cities like Madrid and Barcelona generates the kind of standardized reporting, claims handling, and back-office work that AI tools take on readily. National-level automation pressure therefore looks moderate only because it averages across very different kinds of work.
- Sector: Spain's employment is concentrated in tourism and hospitality, retail, agriculture, construction, and an expanding financial and business-services sector clustered in its largest cities. AI pressure is strongest in administrative and back-office functions: insurance processing, call-center scripts, standardized financial reporting, and routine translation or documentation work. It is weakest in front-line hospitality roles such as hotel staff, waiters, and tour guides, in seasonal agricultural work tied to specific crops and regions, and in construction trades, all of which depend on physical presence, timing, and direct human contact that resist automation.
- Resilience: Spain's durability rests on the sheer scale of its in-person service economy. Hospitality and tourism roles that depend on hosting, language, and cultural familiarity hold their value because visitors pay for human warmth as much as efficiency, and seasonal agricultural work remains tied to physical harvesting that machines and software cannot substitute for at scale. Construction and skilled trades, driven by ongoing housing and infrastructure activity, also stay resistant because they require site-specific judgment and manual skill rather than desk-based processing.
- Limits: A national score obscures the sharp seasonal swings in Spain's labor market, where tourism and agricultural employment expand and contract with the calendar, and it also flattens the difference between coastal, tourism-heavy regions and Spain's more industrial or rural interior. Read the score as an average across a labor market where the timing of the year and the region in question can matter as much as the sector itself, especially for workers whose jobs exist only part of the year.

### Sweden

- URL: https://ai-job-risk.net/countries/sweden
- Lead: Sweden combines a strong engineering and industrial base with a vibrant technology and startup sector, a very large public sector, and labor-market institutions built around strong unions and collective bargaining. AI exposure concentrates in administrative and standardized technical work, while Sweden's emphasis on lifelong learning, strong worker protections, and a public sector oriented toward retraining means displacement pressure is often absorbed through negotiated transition rather than abrupt job loss.
- Overview: Sweden is best read through the lens of its institutional strength as much as its industry mix. The country's engineering firms and technology startups generate real AI exposure in software-adjacent and process-engineering roles, but Sweden's collective-bargaining system and active labor-market policies mean the human impact of that exposure is shaped as much by negotiated retraining and transition support as by the technology itself. Its very large public sector, covering healthcare, education, and elder care, also anchors substantial employment in relationship-based work that AI can support but not replace.
- Sector: Sweden's employment concentrates in engineering and industrial manufacturing, a globally competitive technology and startup sector, telecom and forestry-linked industries, and an unusually large public sector delivering healthcare, education, and social care. AI pressure is strongest in standardized engineering documentation, software testing, financial administration, and back-office functions inside both industrial firms and government agencies. It is weaker in frontline public-sector roles such as nursing, elder care, and teaching, and in the hands-on engineering and manufacturing work that still requires physical problem-solving on production lines and in the forestry and mining operations that anchor Sweden's industrial regions.
- Resilience: Sweden's resilience comes from the combination of strong worker protections and a public sector built around direct human care. Healthcare workers, teachers, and elder-care staff remain essential because their work is relational and physically present in ways AI cannot substitute for, and Sweden's collective-bargaining culture gives displaced workers structured pathways into retraining rather than leaving them exposed. Skilled engineers and technicians in advanced manufacturing also hold their value because Swedish industry competes on precision and innovation rather than low-cost repetition.
- Limits: A national score cannot capture how Sweden's strong labor-market institutions change the practical meaning of exposure: a role can be technically automatable while remaining protected by collective agreements, active retraining programs, and a policy tradition that treats labor transitions as a shared responsibility. Read the score as a technical exposure measure, not a prediction of job loss, in a country where institutions actively mediate that transition.

### Switzerland

- URL: https://ai-job-risk.net/countries/switzerland
- Lead: Switzerland's economy is anchored by global private banking and financial services, a world-leading pharmaceutical and life-sciences industry, and precision manufacturing in instruments, watches, and machinery, all layered with a vocational training system that keeps technical expertise deeply embedded in the workforce. AI exposure concentrates in financial back-office processing and standardized compliance work, while pharmaceutical research, precision manufacturing craftsmanship, and Switzerland's high-value advisory services remain anchored in specialized expertise that is not easily automated.
- Overview: Switzerland is best read as an economy built on high-value specialization rather than volume, which changes how AI exposure plays out. Its banking sector generates standardized compliance, reporting, and client-administration work well suited to AI assistance, but its wealth-management core depends on discretion and regulatory judgment that resist automation. Its pharmaceutical companies run advanced research operations, where AI speeds up data analysis but scientific judgment and regulatory approval remain human-led. Its precision manufacturing sector, especially watchmaking and medical instruments, depends on craftsmanship and exacting tolerances that keep production work firmly hands-on.
- Sector: Switzerland concentrates employment in private banking and asset management, pharmaceuticals and life sciences, precision manufacturing including watches and medical devices, and a vocational-education-linked industrial base. AI pressure is strongest in banking compliance, standardized reporting, insurance administration, and routine legal and financial documentation. It is markedly weaker in wealth-management advisory roles that depend on client trust and discretion, in pharmaceutical research and regulatory affairs that require scientific judgment, and in precision manufacturing roles such as watchmakers, machinists, and quality inspectors, whose work depends on manual dexterity and tolerances measured in fractions of a millimeter.
- Resilience: Switzerland's durability rests on specialization that took decades to build and cannot be easily replicated by general-purpose AI. Wealth-management advisors, pharmaceutical scientists, and precision manufacturing craftsmen all depend on forms of expertise, discretion, or manual skill developed through Switzerland's strong vocational and apprenticeship system, which keeps technical knowledge embedded in people rather than fully codified in documents. This apprenticeship tradition also means Switzerland continually renews its supply of skilled trades workers whose value lies in hands-on precision.
- Limits: A single national score understates how much Switzerland's labor market is split between a small number of extremely high-value, judgment-intensive sectors and more standardized administrative layers that support them. Read this score with that split in mind, since a compliance analyst at a private bank and a research scientist at a pharmaceutical firm face very different pressure despite sharing the same national score and the same small, wealthy economy.

### United Arab Emirates

- URL: https://ai-job-risk.net/countries/united-arab-emirates
- Lead: The UAE's labor market runs on a large expatriate workforce spread across finance, logistics, tourism, real estate, and a shrinking but still significant oil sector, all under a government that has explicitly made AI adoption a national policy priority. That top-down push means AI tools reach white-collar and administrative work unusually fast, while the country's role as a logistics and trade hub, along with its hospitality and construction-driven real estate sector, keeps a large share of physically grounded work intact.
- Overview: The UAE is easiest to read by separating its government-directed digital economy from the physical logistics, hospitality, and construction work that keeps the country running. Free-zone finance firms, corporate headquarters, and government agencies have moved quickly to adopt AI in administration, reporting, and customer service, following explicit national strategy. Meanwhile the ports, airports, hotels, and construction sites that make Dubai and Abu Dhabi function depend on an expatriate labor force doing work that is physically located and situationally variable, which changes far more slowly regardless of policy ambition.
- Sector: The UAE concentrates employment in trade and logistics, financial services, tourism and hospitality, real estate and construction, and government administration, with oil and gas still economically significant but employing a comparatively small, specialized workforce. AI pressure is strongest in banking back-office functions, corporate administration, government service processing, and customer support, areas the national AI strategy explicitly targets. It is much lighter in port and airport logistics operations, hotel and hospitality service roles, and construction, where expatriate labor performs physical, on-site work that automation does not reach.
- Resilience: What stays durable in the UAE is work tied to physical logistics, hospitality service, and construction execution. Port and airport operations depend on people managing physical cargo flows and exceptions in real time, hotel and tourism roles depend on personal service that visitors specifically pay for, and the country's continuous construction and real-estate development requires on-site trades that cannot be virtualized. This physical layer, largely staffed by expatriate workers, sits apart from the government's AI-adoption push aimed at the administrative and financial layer above it.
- Limits: A single UAE score has to reconcile a government actively accelerating AI adoption in administration and finance with a physical logistics, hospitality, and construction economy that changes on its own slower timeline. The country's heavy reliance on expatriate labor across skill levels also means exposure varies enormously by visa category and sector, something a national average cannot show. Read the score alongside whether the work in question sits inside the policy-driven digital economy or the physical trade and hospitality economy.

### United Kingdom

- URL: https://ai-job-risk.net/countries/united-kingdom
- Lead: The United Kingdom's labor market leans heavily on financial and professional services centered in and around London, alongside a large public sector, notably the NHS, and a distinctive creative industries cluster spanning media, advertising, gaming, and design. That mix puts an unusually large share of employment into contract-heavy, document-intensive knowledge work exposed to AI support, while regulated professions, healthcare delivery, and creative origination retain forms of judgment and accountability that are harder to compress into automated output.
- Overview: The UK is best read by separating the financial and professional services core from the public and creative sectors around it. Legal drafting, audit support, insurance processing, and standardized financial analysis sit in the exposed layer, since the City of London runs on exactly the kind of structured, precedent-based document work that generative tools handle well. Frontline NHS care, regulated legal advice, teaching, and creative direction sit in the more durable layer, where statutory accountability, direct patient contact, or original creative judgment keep human involvement central even as supporting tasks get faster.
- Sector: Employment concentrates in financial services, professional and business services, healthcare and social care, education, and a creative and media sector that is large relative to the size of the economy. AI pressure is strongest in back-office banking, insurance claims handling, paralegal work, accounting, and standardized market research, all clustered around London and other financial centers. It moves more slowly through NHS nursing and clinical care, skilled trades, and creative roles such as directing, writing, and design, where professional regulation, patient safety obligations, and the premium on original creative work limit how far automation can go.
- Resilience: The most durable work combines regulatory accountability with human judgment: solicitors and barristers bound by professional conduct rules, NHS clinicians accountable for patient outcomes, and creative professionals whose value lies in original work rather than reproducible output. Britain's strong tradition of professional self-regulation across law, medicine, and accountancy also means licensing bodies, not just employers, control how quickly AI tools can substitute for supervised human sign-off, keeping adoption gradual even where the underlying tasks are technically automatable.
- Limits: A UK-wide score compresses a sharp divide between London's finance and professional-services economy and the rest of the country, where manufacturing, retail, and public-sector employment dominate and AI exposure looks quite different. It also cannot separate roles where AI speeds up drafting and analysis from roles where statutory and professional accountability still requires a named, liable human. Read the score alongside sector concentration and regional variation, not as one uniform verdict.

### United States

- URL: https://ai-job-risk.net/countries/united-states
- Lead: The United States runs on a uniquely large layer of white-collar knowledge work — software, finance, insurance underwriting, corporate law, consulting, and administrative support — sitting alongside at-will employment rules that let firms restructure headcount faster than in most other economies. That combination makes AI adoption both unusually visible and unusually fast to convert into actual job changes, while a huge, geographically spread service and care economy, from hospitals to logistics yards, keeps a large share of work tied to physical presence and direct human contact.
- Overview: Reading the United States well means separating the coastal, corporate knowledge economy from the service and operational economy that employs most workers day to day. Drafting, analysis, coding support, customer correspondence, and standardized financial or legal document work sit in the exposed layer, since employers face few contractual barriers to reorganizing these roles quickly. Healthcare delivery, skilled trades, logistics, retail floor work, and public-sector jobs sit in the durable layer, where physical tasks, licensing, and direct accountability to patients or customers slow substitution regardless of how capable the underlying models become.
- Sector: Employment concentrates in professional and business services, healthcare, retail, finance, technology, and a large logistics and warehousing sector built around national supply chains. AI pressure is sharpest in back-office finance, insurance claims processing, paralegal and contract review work, customer support centers, and entry-level software tasks, all of which are text-heavy and already digitized. It spreads more slowly through nursing and direct patient care, construction trades, truck driving and warehouse operations, and K-12 teaching, where physical work, licensure, and in-person trust remain central and where a fragmented, state-by-state regulatory system slows any uniform rollout.
- Resilience: What holds up best are roles combining licensed responsibility with physical or interpersonal work: registered nurses, electricians and HVAC technicians, teachers, and first responders. The country's decentralized regulatory structure, with fifty separate states setting licensing and liability rules on their own terms, also slows any single wave of standardized automation from moving uniformly through regulated professions like healthcare and law, even as software tools inside those professions keep improving and law firms experiment with drafting assistance.
- Limits: A single national score flattens enormous regional variation, from a Bay Area labor market dense with software roles to manufacturing and agricultural regions where AI exposure looks completely different. It also blends states with strong at-will hiring and firing against those with stronger worker protections. Read the number as a broad signal about a knowledge-heavy, flexible labor market, not as a claim that Silicon Valley and small-town Ohio face the same transition.