59% of hiring managers say they deliberately overstated artificial intelligence as the reason for their layoffs, according to a ResumeBuilder survey of 1,000 managers published in March1. Only 9% said AI had actually replaced roles at their organisations. ResumeBuilder is a US job-search and CV platform; the survey asked managers directly why they had cut staff and whether AI was the genuine cause. The question turned live again this week, as the Office for National Statistics reported UK unemployment at 4.9% and the European Parliament stripped a worker AI-literacy right.
The 59% reads, on its own, as proof the job losses are exaggerated. Oxford Economics, a UK forecasting firm, puts genuine AI-driven cuts at around 4.5% of US layoffs in the first 11 months of 2025, roughly 55,000 positions against about 245,000 lost to ordinary cost pressure. Firms credit the machine for decisions the budget had already made. Sam Altman of OpenAI has acknowledged the practice, the same week his company moved toward a September listing above one trillion dollars.
The bigger number runs the other way. The Stanford Digital Economy Lab, Stanford University's economics research unit under Erik Brynjolfsson, calculates that AI is preventing roughly 950,000 to one million US hires a year. Set that against about 145,000 declared AI layoffs since 2023, per the outplacement firm Challenger, Gray & Christmas through May. Most of the loss never reaches a redundancy notice; it shows up as the requisition never opened and the graduate never called back, so a graduate entering the workforce this year faces a far tighter market than one who arrived in 2023. MIT Sloan economist Paul Osterman told Fortune in May that AI attribution often dresses up cuts that were planned anyway, and the New York Federal Reserve has found displacement signals that predate ChatGPT entirely.
