Jobs Already Being Replaced by AI — and the Jobs Being Displaced Now
Some jobs are not waiting for a distant future to feel the impact of AI. In translation, customer support, back-office clerical work, and commercial writing, generative AI and automated agents are already absorbing tasks that used to fill a full working day. This page collects the occupations where that shift is clearest today, separates the ones where replacement is already well advanced from the ones where the pressure is building fast, and pairs each with our own live AI risk score so you can see direction as well as evidence.
Edition: 2026-Q2Last updated: June 1, 2026Current AI risk score
9Jobs where replacement is already well advanced
7Jobs being displaced right now
16Sources · 5+
What "replaced by AI" really 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.
How to read this page
Each entry combines two things: durable, sourced evidence that we revise about once a quarter, and a current AI risk score drawn live from our weekly index. The prose explains what changed and what still needs people; the score badge shows where the job sits right now and how it moved week over week. Read them together, follow the link to the full job profile for the detailed breakdown, and treat the cited figures as snapshots from their original reports rather than guarantees about any individual role.
For these roles, there is documented evidence that AI is already absorbing the core tasks at scale. The work is not gone, but its routine center is shrinking fast.
A1
Translators and interpreters
73AI
+0 vs last week
What AI changed
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.
Evidence
A 2024 Society of Authors survey found more than a third of translators had already lost work to generative AI, and an Oxford analysis cited by CNN in 2026 estimated that heavier machine-translation adoption was associated with tens of thousands fewer translator roles. Duolingo cut roughly 10% of its contractors in early 2024 as it moved content generation to GPT-4.
What still needs people
Literary, legal, medical, and diplomatic translation still depend on people who carry nuance, cultural context, and legal responsibility. Live interpretation under pressure, where a mistranslation has real consequences, remains a human-led task.
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.
Evidence
In February 2024 Klarna said its OpenAI-powered assistant was doing the work of about 700 full-time agents and handling two-thirds of service chats. The figure was cost-avoidance rather than 700 layoffs, and by 2025 Klarna began re-recruiting humans for quality reasons — a useful reminder that full automation can overshoot. Earlier, in 2023, e-commerce firm Dukaan said it cut around 90% of its support staff after deploying a chatbot, a move that drew heavy criticism.
What still needs people
Complex complaints, emotionally charged situations, edge cases, and anything requiring real authority to make an exception still route to people. The role is shifting from high-volume first response toward supervising AI and handling what it cannot.
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.
Evidence
The World Economic Forum's Future of Jobs Report 2025 names clerical and secretarial roles — data-entry clerks, administrative and executive secretaries, and bank tellers — as the largest declining occupations in absolute terms. In 2023 IBM's CEO said the company would pause hiring for around 7,800 back-office roles it expected AI to automate over time.
What still needs people
Coordination that spans teams, judgement about unusual cases, and the relationship-management side of administrative work are far harder to standardize and tend to stay with people even as the keystroke-heavy tasks fall away.
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.
Evidence
A widely cited study of freelance platforms found demand for automatable writing work dropped sharply after ChatGPT's release, with some writing categories down around 30%. Reporting through 2024 collected accounts of writers asked to use AI tools until their roles were cut.
What still needs people
Original reporting, a distinctive brand voice, persuasive work that depends on understanding a specific audience, and anything where accuracy and reputation are on the line still reward skilled human writers — increasingly as editors and directors of AI output.
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.
Evidence
The U.S. Bureau of Labor Statistics projects employment of bookkeeping, accounting, and auditing clerks to decline about 6% from 2024 to 2034, explicitly attributing the fall to automation. This is the clerk-level role specifically — licensed accountants and auditors, whose work centers on judgement and assurance, have a different, more stable outlook.
What still needs people
Interpreting unusual transactions, advising on decisions, and taking responsibility for what the numbers mean stay with people. Automation removes the data-shuffling, not the accountability.
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.
Evidence
The BLS Occupational Outlook Handbook projects only about 1% growth for editors through 2034 — slower than average — and points to productivity tools as a drag on demand. The defensible reading is suppressed growth, not a sudden collapse in head-count.
What still needs people
Substantive editing — structure, argument, tone, fact-checking, and protecting a publication's credibility — depends on judgement that mechanical correction tools do not provide.
AI voice agents now handle outbound calling, lead qualification, and routine follow-ups, automating the scripted, high-repetition core of the role.
Evidence
Telemarketing is consistently ranked among the occupations at highest automation exposure in workforce analyses such as Goldman Sachs', and AI-voice calling tools have become widely available. In February 2024 the U.S. FCC ruled that AI-generated voices in robocalls are "artificial" under existing law, a sign of how fast the technology spread into outbound calling.
What still needs people
Complex, consultative selling that depends on reading a person and building trust over time is much harder to automate than scripted outbound dialing.
Text-to-image generation produces usable visuals on demand, undercutting commissions for routine, high-volume illustration and template-style design work.
Evidence
A 2024 Society of Authors survey found about a quarter of illustrators had already lost work to AI, and many reported the value of illustrated work falling. Job-board tracking through 2024 and 2025 showed steep declines in graphic-art postings.
What still needs people
Art direction, brand identity, conceptual work, and design that has to solve a specific business problem still reward people who can decide what is worth making — not just generate more options.
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.
Evidence
Clio's 2024 Legal Trends Report estimated that a large majority of hourly paralegal tasks are automatable by AI. Notably, BLS still projects modest growth for paralegals, so this is best read as the role being reshaped task-by-task rather than eliminated outright.
What still needs people
Client interaction, supervising and verifying AI output, courtroom and filing logistics, and the parts of legal work that carry professional responsibility remain firmly human.
For these roles, the pressure is building quickly. AI is automating key tasks and thinning entry-level openings, even if large-scale displacement is still emerging.
B1
Junior software developers
62AI
+1 vs last week
What AI changed
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.
Evidence
A 2025 Stanford Digital Economy Lab study found employment of software developers aged 22–25 fell roughly 20% from its late-2022 peak, even as older developers held steady. Some firms have publicly slowed or paused junior engineering hiring, citing AI productivity gains.
What still needs people
System design, judgement about what to build and what to refuse, debugging under uncertainty, and accountability for production systems are where experienced engineers keep their value — and senior demand has held up better than junior.
Generative AI now drafts research summaries, standardizes reporting, and pulls together comparisons, automating the routine synthesis layer of analyst work.
Evidence
The WEF Future of Jobs Report 2025 and broad exposure estimates from Goldman Sachs place mid-level cognitive analysis among the white-collar tasks most exposed to generative AI, though documented head-count cuts are still thinner than in the Tier A roles above.
What still needs people
Framing the right question, designing sound methodology, and judging which findings actually matter to a decision are harder to hand to a model than the write-up of results.
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.
Evidence
Goldman Sachs has piloted AI for producing pitchbooks, reportedly cutting the routine hours involved, and the firm's own research places business and financial operations among the more automatable task categories. These are productivity shifts so far rather than mass layoffs, but they change how many juniors a desk needs.
What still needs people
Judgement under uncertainty, client relationships, deal negotiation, and ownership of recommendations stay with people; the model accelerates the analysis, it does not sign off on it.
AI agents now source candidates, screen applications, and schedule interviews, automating large parts of the coordination-heavy core of recruiting.
Evidence
Industry surveys through 2025 report that the overwhelming majority of employers now use AI somewhere in hiring, with automated scheduling and screening cutting coordination time substantially. So far this reads more as heavy task automation than documented mass displacement.
What still needs people
Persuading scarce candidates, assessing fit and judgement, closing offers, and the relationship and advisory side of hiring remain people-led — recruiters shift toward strategy as the admin is automated.
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.
Evidence
Analyses by Semrush and Ahrefs through 2024–2025 found Google's AI Overviews cut organic click-through rates substantially on informational queries. The pressure shows up first as falling traffic that undercuts demand, rather than as direct head-count cuts, and the field is reorienting toward visibility inside AI answers.
What still needs people
Strategy, brand positioning, original research and data, and earning genuine authority are harder to automate than keyword-driven content production.
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.
Evidence
U.S. journalism shed thousands of jobs in 2024, and some newspaper chains have publicly moved toward AI-created articles. Importantly, a 2026 study found the contraction was driven mainly by the collapse of digital-ad and traffic economics — worsened by AI search — rather than AI writing replacing reporters one-for-one.
What still needs people
Original reporting, source relationships, verification, and accountability for what gets published are the core of journalism and are not something a model can take responsibility for.
AI voice synthesis now narrates audiobooks, ads, and explainer content, taking on routine narration work that used to be recorded by people.
Evidence
Apple Books and Google Play have launched AI-generated narration, and Audible has introduced ways for narrators to clone their own voices. Industry data reports estimate AI narration reached a meaningful share of new audiobook releases in 2025.
What still needs people
Character-rich performance, emotional range, and work where a specific human voice is part of the value still belong to performers — and consent and licensing of one's own voice are becoming central issues.
It is easy to read a page like this as a forecast that these jobs will vanish — but the evidence points to something more specific and less total. AI is absorbing tasks, not whole professions, and it tends to hit the repetitive, well-defined, entry-level parts of a job first. The work that survives concentrates on judgement, accountability, relationships, and handling the exceptions a model gets wrong. Some of the boldest corporate claims have also been walked back: Klarna began re-hiring human agents after over-automating support. New roles are appearing too — the WEF projects far more jobs created than displaced by 2030, even as the mix shifts hard. The useful takeaway is not that any one job is doomed, but that the routine core of many jobs is being automated, and the people who stay valuable are the ones who move toward the parts machines still cannot own.
We update this report about once a quarter using a mix of our own weekly risk data, labor-market reporting, and primary research. The risk score on each card is pulled live from our index and changes every week; the written analysis is revised on the quarterly cycle. Use the links to read each job's full profile, and treat the cited statistics as evidence from their original sources rather than predictions about your specific situation.
Risk scores are pulled live from our weekly AI Job Risk Index and change every week. The written analysis is reviewed about once a quarter. Cited figures are snapshots from their original sources, not predictions about any individual job.
