Updated: Washington, May 6, 2026, 08:00 EDT
The answer to the headline question “How many jobs have been lost to AI in 2023?” is not millions. Using Challenger, Gray & Christmas data, U.S. employers linked 4,247 announced job cuts to artificial intelligence in 2023, based on the firm’s January 2024 figure of 4,628 AI-linked cuts since May 2023, including 381 that January; Andrew Challenger said companies were “not outright blaming AI” for many layoff decisions. Challenger Gray & Christmas
The issue matters now because that small early count has turned into a live labor-market signal. Reuters reported that AI accounted for 54,694 planned U.S. layoffs in the first 11 months of 2025, while announced hiring plans fell to their lowest year-to-date level since 2010.
By March 2026, Challenger said AI led all stated reasons for U.S. job cuts that month, with 15,341 announced cuts, or 25% of the total. The firm said AI ranked fifth for the first quarter, with 27,645 cuts, and cited technology peers including Dell, Oracle and Meta as companies in a sector shifting spending toward artificial intelligence. “Companies are shifting budgets toward AI investments at the expense of jobs,” Andy Challenger said. Challenger Gray & Christmas
The broader labor market has not broken. The Labor Department’s Job Openings and Labor Turnover Survey, known as JOLTS, showed 6.866 million job openings in March 2026, 5.554 million hires and a layoffs-and-discharges rate of 1.2%; JOLTS measures labor demand and separations, but it does not say whether AI caused a job to disappear.
That gap is the story. The 2023 AI layoff number is a count of announced U.S. cuts where employers supplied AI as a reason, not a global census of every job changed, frozen, outsourced or quietly thinned by software.
Generative AI — systems that create text, code, images or other content from prompts — has always looked bigger in forecasts than in layoff notices. Goldman Sachs economists Joseph Briggs and Devesh Kodnani wrote in 2023 that AI could expose the equivalent of 300 million full-time jobs to automation, but also said most exposed jobs were more likely to be complemented than replaced.
Some early workplace evidence points the same way. NBER summarized research by Erik Brynjolfsson, Danielle Li and Lindsey Raymond showing customer-support agents using an AI tool resolved 13.8% more issues per hour, with the largest gains among less experienced workers.
The risk is that layoffs may be the late indicator. If employers use AI to slow entry-level hiring, shrink graduate programs, replace contractors or avoid backfilling routine roles, the market can weaken without a clean “AI layoff” tag. The downside case is not just mass firings; it is fewer first jobs and a thinner path into white-collar work.
The OECD has also warned against reading the early data too calmly. It found little evidence so far of a major negative employment effect from AI, but said 27% of jobs across member countries were in occupations at high risk of automation; Secretary General Mathias Cormann said the outcome would depend on policy action.
For 2023, the clean answer is about 4,247 announced U.S. job cuts tied directly to AI. For 2026, the better answer is less tidy: the number is rising, the causes are mixed, and the first damage may show up in hiring plans before it appears in pink slips.
