Paul Osterman, a labour economist at the MIT Sloan School of Management, told Fortune that AI attribution in layoff announcements is largely a cover story for cuts firms had already planned1. "They've been saying that for 20 years," he said of bosses reaching for a technology excuse2. Harvard Business Review research supports him: only 2% of layoffs at AI-citing companies followed an actual AI deployment.
Two points cut the other way. The declared AI share of stated layoff reasons jumped from under 8% in 2025 to 40% in May, faster than a gradual rebranding of routine cuts would move. And Stanford's Digital Economy Lab put the real AI labour impact at roughly 34 times Challenger's declared count, derived from a collapsed hiring rate in the Job Openings and Labor Turnover Survey (JOLTS) rather than from announced firings. The same lab documented the under-25 employment pattern that the youth figures now confirm.
Osterman's point is that the stated reason is unreliable, which can mean the named number overstates real AI displacement or that firms underreport it to avoid scrutiny. Stanford's reading treats the record 38,579 as a floor, the visible edge of a larger effect running through hires never made. Both can hold: the pattern of cutting on record revenue, repeated by Dell, HP Inc and CrowdStrike in late May, shows firms shedding staff for reasons the single stated cause does not fully capture. The argument is over whether the headline AI figure is too high or too low, not over whether the displacement is happening.
