The AI job losses haven't come, but fewer young people are getting hired into jobs AI does well.
71% of Americans expect AI to mean fewer jobs. So far the numbers disagree: unemployment is 4.1%, and studies find no clear link between AI use and widespread job losses. The exception is young people. In the jobs most exposed to AI, employment of 22- to 25-year-olds is 19% lower than it would be had it kept pace with less exposed jobs, though the evidence does not pinpoint AI as the cause of the gap.
Most Americans expect AI to take jobs. In a Pew survey in June, 71% said it would lead to fewer jobs in the United States over the next two decades, up from 64% in 2024. Only 5% expected more1. Workers are also increasingly worried about their own jobs. In August, 27% of US workers told Gallup they fear technology could make their own job obsolete, a record and double the 2017 figure2.
The fear has backing inside the AI industry itself. In May 2025 Dario Amodei, chief executive of Anthropic, said AI could push unemployment to 10 to 20 percent in the next one to five years3,4.
Employers feed the fear too. They have named AI as the reason for 121,401 announced job cuts so far this year, about a fifth of the total5. Some say plainly that AI now does the work. Salesforce cut about 4,000 customer-support roles, with its chief executive saying the company needed “less heads” because AI now handles the work. Oracle told regulators that “the adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce”6.
The worry extends to the professions. An MIT mathematician wondered whether mathematicians are “just the canaries in the coal mine”7. In a survey of professional writers who chose to take part, “Nearly half of freelancers have experienced reduced demand for their work due to AI”8.
Some have gone beyond words. Hollywood’s writers and actors both struck in 2023, with AI among the issues. The writers’ new contract says “AI can’t write or rewrite literary material”, while the actors’ contract requires their consent for most uses of a digital copy of them9,10. Video-game performers later struck for almost a year over AI alone, winning the right to consent before AI uses their voice and likeness11. More than 4,000 mathematicians have signed the Leiden Declaration, endorsed by the International Mathematical Union, warning that AI may disrupt the field’s “mechanisms for hiring, funding, and recognition”12.
No sign in the job numbers
Across the economy, AI isn’t showing up in the job numbers. US unemployment was 4.1% in October13. Announced job cuts are down 44% on last year, and in the autumn “Market/Economic Conditions” overtook AI as the reason employers most often gave for cuts5.
In October, Yale’s Budget Lab found that “Measures of AI usage show no connection to changes in employment or unemployment”14. The OECD sees “no signs of widespread labour displacement”15. The Dallas Fed estimates that even if every job lost among young workers in the jobs where AI can do the most tasks had turned into unemployment, it would add only 0.1 percentage point to the national rate16.
Oxford Economics offers a simple test: “If AI were already replacing labour at scale, productivity growth should be accelerating.” So far, it says, “firms don’t appear to be replacing workers with AI on a significant scale”17.
Some jobs often described as doomed are growing. The number of US software developers rose about 10% from 2022 to 202518. The number of US translators, often named as AI’s first casualty, was flat over the same years18. Freelance platforms tell a different story: after ChatGPT arrived, demand for writing and translation there fell 20–50% below where it would otherwise have been19.
Forecasts have been wrong before. In 2016 Geoffrey Hinton suggested that training radiologists was no longer worth it20. Since then the number of US radiologists has grown, and US radiology training programs filled 95–98% of their places in 202621,22. As Gallup puts it: “So far, workers’ growing concern appears to be outpacing actual job displacement due to AI”2.
Fewer young hires
The picture is less reassuring for young workers. Amodei’s warning had a second half: AI could “wipe out half of all entry-level white-collar jobs”3. Several studies point to a pattern consistent with that concern: young people are being hired less into the jobs AI does best.
Using ADP payroll data, Stanford researchers find that employment of 22- to 25-year-olds in the jobs most exposed to AI is 19% lower than it would be had it kept pace with less exposed jobs. Older workers show no such gap. The gap “operates primarily through reduced hiring of young workers rather than increased separations”23.
Other studies point the same way. In US résumé data, junior employment at firms that adopted AI fell about 9% compared with similar firms that had not24. In a New York Fed survey, about 12% of service firms using AI said they had hired fewer people because of it; 1% said they had let workers go25.
Software is the clearest case. Entry-level jobs were 4.5% of software development postings on Indeed in early 2026; senior roles were 69.3%26. Across 41 countries, firms adopting AI cut the junior share of their staff, mostly by adding senior staff. At those firms’ offices abroad, AI “changes which workers an affiliate employs more than how many”27.
People inside the professions report the same. A declaration by mathematicians warns that AI’s effects fall “disproportionately” on “students and early-career mathematicians”28. A computer scientist wrote in Science that colleagues “are quietly refraining from taking on as many students as they used to”, and that he “probably wouldn’t recruit my younger self”29.
Officials see warning signs too. Federal Reserve governor Michael Barr said AI “may be starting to adversely affect some groups, in particular young people who are just starting their careers in some sectors”30. The Reserve Bank of Australia sees “little evidence of economy-wide effects so far, but some signs of weaker employment and job-finding rates – rather than higher layoffs – for entry-level or younger workers in AI-exposed occupations”31.
But there are other plausible explanations.
The first is timing: the trouble started before ChatGPT. The risk of unemployment in jobs exposed to AI began rising in early 2022, with “gaps opening before late 2022”32. Stanford itself notes “some divergent trends predating generative AI”23.
Young people in AI-exposed jobs are also not the only ones struggling. Teen unemployment rose 4 percentage points from its low, compared with 1.5 points for recent graduates33. Young people without degrees did at least as badly34.
Higher interest rates offer another explanation. Two Google economists, whose employer sells AI, call the pattern “the predictable consequences of a classic macroeconomic shock: the sharpest monetary policy tightening cycle in four decades”. They argue that jobs exposed to AI are concentrated in the industries hit hardest by higher rates35. In Sweden, the fall in job postings since 2022 “aligns with monetary tightening rather than AI”36.
Remote work may also explain 64% of the rise in young graduates’ unemployment through 202437. And there are many more computer science graduates competing for work: the field’s jobless rate “reflects in part the huge numbers of students who enrolled during the coding boom”38.
The studies themselves disagree. In Denmark, “the declines are not driven by firms adopting AI chatbots”4, while the US résumé study links the declines to AI adoption24.
Meanwhile, hiring has become harder for everyone. Applications per job more than doubled from 2022 to 2025, while recruiters per firm fell 56%39. AI-written cover letters stopped telling employers much40,41.
Our take
So far, AI is not taking jobs on a scale that shows up in the job numbers. Unemployment is 4.1%13, and studies find no clear link between AI use and widespread job losses14,15. Forecasts of mass job loss may yet come true, but they have not so far.
The one place AI may be costing jobs is at entry level: young people are being hired less into the jobs AI does best. Derek Thompson made the case in The Atlantic in April 202542.
Why “may”? The pattern is narrow and clearest in US software, and the evidence does not pinpoint AI as the cause. Nor is it the same everywhere: the drop in the share of junior staff is significant in the US and UK27, while studies in Denmark and Finland find no such effect4,43.
Two things can be true. AI helps novices most: among customer-support agents, less skilled and less experienced workers resolved 30% more issues per hour with AI help44. Yet fewer novices are being hired.
If the trend continues, fewer juniors now could mean fewer trained seniors later. “If they don’t change how hiring is currently happening, they’re not going to have mid-level career people in a few years,” says Tracy Camp of the Computing Research Association38. Barr warns that “for these workers, the short run may have long-term consequences”30.
We will change our view if Yale’s tracker begins to link AI use to job losses14, or if Stanford’s monthly updates show the gap spreading beyond young workers23.
Sources
- Colleen McClain & Eugenie Park, "Young US adults are increasingly wary of AI, concerned it will take jobs" (Pew Research Center, short read). August 18, 2026. pewresearch.org/short-reads/2026/08/18/young-adults-in-the-us-are-increasingly-wary-of-ai-concerned-it-will-take-jobs/
- More U.S. Workers Fear Losing Their Jobs to Technology (Gallup, Work and Education survey). September 15, 2026. news.gallup.com/poll/714368/workers-fear-job-losses-technology.aspx
- America Isn’t Ready for What AI Will Do to Jobs (The Atlantic, March 2026 issue). Josh Tyrangiel, The Atlantic. 10 February 2026. theatlantic.com/magazine/2026/03/ai-economy-labor-market-transformation/685731/
- Still Waters, Rapid Currents: Early Labor Market Transformation under Generative AI. Anders Humlum, Emilie Vestergaard. March 2026. rfberlin.com/wp-content/uploads/2026/03/26078.pdf
- Challenger Report, October 2026 (job-cut announcements). NOVEMBER 5, 2026. challengergray.com/wp-content/uploads/2026/11/Challenger-Report-October-2026.pdf
- Monday.com is the latest tech company to blame AI for layoffs — here are 20 others (TechCrunch running list, "major tech layoffs in 2026 where employers cited AI"). June 22, 2026. techcrunch.com/2026/06/22/the-running-list-major-tech-layoffs-in-2026-where-employers-cited-ai
- AI triggers a wave of anxiety among the world's greatest math minds (The Washington Post). Miriam Waldvogel, The Washington Post. 19 August 2026. thenightly.com.au/business/the-washington-post-ai-triggers-a-wave-of-anxiety-among-the-worlds-greatest-math-minds-c-22749535
- A.I. and the Writing Profession: A Comprehensive Survey & Analysis (Gotham Ghostwriters / Josh Bernoff). Nov 2025. gothamghostwriters.com/wp-content/uploads/2025/11/AI-Writer-Survey.pdf
- Summary of the 2023 WGA MBA (Writers Guild of America, contract campaign site). 9 Oct 2023. wgacontract2023.org/the-campaign/summary-of-the-2023-wga-mba
- TV/Theatrical Contracts 2023: Summary of Tentative Agreement (SAG-AFTRA). 16 Nov 2023. sagaftra.org/sites/default/files/sa_documents/TV-Theatrical_23_Summary_Agreement_Final.pdf
- SAG-AFTRA Members Ratify Interactive Media Agreement After 320-Day Video Game Strike (TheWrap). Jeremy Fuster, TheWrap. 9 Jul 2025. thewrap.com/sag-aftra-ratify-interactive-media-agreement-video-game-strike/
- The Leiden Declaration on Artificial Intelligence and Mathematics. 16 researchers from 15 universities. 2 June 2026. leidendeclaration.ai/
- The Employment Situation — October 2026 (BLS). U.S. Bureau of Labor Statistics. November 6, 2026. bls.gov/news.release/empsit.nr0.htm
- Tracking the Impact of AI on the Labor Market (October 2026 update). The Budget Lab at Yale. October 14, 2026. budgetlab.yale.edu/research/tracking-impact-ai-labor-market
- OECD Economic Outlook, Volume 2026 Issue 1 ("Under Pressure"), June 2026 — labour-market box. oecd.org/content/dam/oecd/en/publications/reports/2026/06/oecd-economic-outlook-volume-2026-issue-1_8be0dba6/2d1956f0-en.pdf
- Young workers' employment drops in occupations with high AI exposure. Tyler Atkinson and Shane Yamco. January 06, 2026. dallasfed.org/research/economics/2026/0106
- Oxford Economics, "Evidence of an AI-driven shakeup of job markets is patchy" (Research Briefing). 7 Jan 2026. oxfordeconomics.com/resource/evidence-of-an-ai-driven-shakeup-of-job-markets-is-patchy/
- Occupational Employment and Wage Statistics (OEWS): computer programmers, software developers, customer service representatives, interpreters and translators. U.S. Bureau of Labor Statistics. May 2022. bls.gov/oes/2022/may/oes151251.htm
- Winners and losers of generative AI: Early evidence of shifts in freelancer demand (Teutloff et al.). Ole Teutloff, Johanna Einsiedler, Otto Kässi, Fabian Braesemann, Pamela Mishkin. doi.org/10.1016/j.jebo.2024.106845
- Your A.I. Radiologist Will Not Be With You Soon (NYT). Steve Lohr, The New York Times. 14 May 2025. nytimes.com/2025/05/14/technology/ai-jobs-radiologists-mayo-clinic.html
- The Radiologist Shortage: A Workforce Update from HPI (ACR Bulletin). Elizabeth Y. Rula, PhD. 5 Feb 2026. acr.org/Clinical-Resources/Publications-and-Research/ACR-Bulletin/2026/radiologist-shortage-work-force-update
- Results and Data: 2026 Main Residency Match (NRMP). National Resident Matching Program. May 2026. nrmp.org/wp-content/uploads/2026/05/Main_Match_Results_and_Data-2026.pdf
- Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence (August 2026 revision). Erik Brynjolfsson, Bharat Chandar, Ruyu Chen. August 12, 2026. digitaleconomy.stanford.edu/app/uploads/2026/08/Canaries_August2026.pdf
- Generative AI as Seniority-Biased Technological Change: Evidence from U.S. Résumé and Job Posting Data. Seyed M. Hosseini, Guy Lichtinger. August 31, 2025. raw.githack.com/s-mahdihosseini/Website/b99fa2962dcc4c165bbc3c41dd283afa99c5ccfb/GenAI_SenBTC_May_2026.pdf
- Are Businesses Scaling Back Hiring Due to AI? (Federal Reserve Bank of New York, Liberty Street Economics). Jaison R. Abel, Richard Deitz, Natalia Emanuel, Ben Hyman, Nick Montalbano. 4 Sep 2025. libertystreeteconomics.newyorkfed.org/2025/09/are-businesses-scaling-back-hiring-due-to-ai/
- The Labor Market Is Tilting Toward Seniority. Felix Aidala & Sneha Puri. July 23, 2026. hiringlab.org/2026/07/23/the-labor-market-is-tilting-toward-seniority/
- How Does AI Change Labor Demand? Evidence from 41 Countries (SSRN working paper). Bharat Chandar. 20 September 2026. papers.ssrn.com/sol3/papers.cfm?abstract_id=7498743
- Mathematicians warn of AI threats to profession as industry encroaches (Ars Technica). Jeremy Hsu, Ars Technica. 2 June 2026. arstechnica.com/tech-policy/2026/06/mathematicians-warn-of-ai-threats-to-profession-as-industry-encroaches/
- Ariel Rosenfeld, "Why I may 'hire' AI instead of a graduate student" (Science, Working Life). 12 Mar 2026. science.org/content/article/why-i-may-hire-ai-instead-graduate-student
- Michael S. Barr, speech on artificial intelligence and the labor market (Federal Reserve Board). 17 Feb 2026. federalreserve.gov/newsevents/speech/barr20260217a.htm
- Monitoring Early Effects of AI on the Labour Market (Reserve Bank of Australia, Bulletin/Insights). Mike Major, RBA staff. 9 Nov 2026. rba.gov.au/publications/insights/australian-economy/2026/monitoring-early-effects-of-ai-on-the-labour-market.html
- AI-exposed jobs deteriorated before ChatGPT. Morgan R. Frank, Alireza Javadian Sabet, Lisa Simon, Sarah H. Bana, Renzhe Yu. 5 Jan 2026. arxiv.org/abs/2601.02554
- Smells Like Teen Unemployment. Adam Ozimek. May 28, 2026. agglomerations.eig.org/p/smells-like-teen-unemployment
- Adam Ozimek & Nathan Goldschlag, "AI and Young-adult Jobs: The Real Mystery" (EIG). 5 Mar 2026. eig.org/ai-and-young-adult-jobs-the-real-mystery/
- Looking for the Ladder: Is AI Impacting Entry-level Jobs? (Economic Innovation Group, guest essay). Zanna Iscenko. 14 Jan 2026. agglomerations.eig.org/p/looking-for-the-ladder
- Same Storm, Different Boats: Generative AI and the Age Gradient in Hiring (RATIO Working Paper 388, full paper). Magnus Lodefalk, Lydia Löthman. 16 March 2026. ideas.repec.org/p/hhs/ratioi/0388.html
- Remote Work Leaves Younger Workers Sidelined. Natalia Emanuel, Emma Harrington, Amanda Pallais. June 1, 2026. libertystreeteconomics.newyorkfed.org/2026/06/remote-work-leaves-younger-workers-sidelined/
- Heather Hollingsworth & Nicky Forster, "Computer science grads face a brutal job market as AI reshapes entry-level work" (Associated Press, in the Los Angeles Times). 24 Sep 2026. latimes.com/business/story/2026-09-24/computer-science-grads-face-brutal-job-market-as-ai-reshapes-entry-level-work
- The Hire Standard: hiring benchmarks 2026 (Greenhouse). Mar 2026. greenhouse.com/recruiting-benchmarks
- Cui, Dias & Ye, "Signaling in the Age of AI: Evidence from Cover Letters". Jingyi Cui, Gabriel Dias, Justin Ye. 24 Aug 2026. arxiv.org/html/2509.25054v2
- Galdin & Silbert, "Making Talk Cheap: Generative AI and Labor Market Signaling". Anaïs Galdin, Jesse Silbert. 14 Nov 2025. jesse-silbert.github.io/website/silbert_jmp.pdf
- Derek Thompson, "Something Alarming Is Happening to the Job Market" (The Atlantic). 30 Apr 2025. theatlantic.com/economy/archive/2025/04/job-market-youth/682641/
- AI Has Not Impacted the Youth Labor Market in Finland (ETLA Working Papers No. 135). Antti Kauhanen, Petri Rouvinen. 27 Jan 2026. etla.fi/wp-content/uploads/ETLA-Working-Papers-135.pdf
- Generative AI at Work (Quarterly Journal of Economics 2025). Erik Brynjolfsson, Danielle Li, Lindsey Raymond. February 4, 2025. doi.org/10.1093/qje/qjae044