A sociologist at College of the Holy Cross named Jeffrey Dixon commissioned a survey of 1,250 US workers this summer, asking a question few people have solid data on: has AI actually cost you a job. Only 3% said yes since 2023. About 6% said they landed a job that did not exist before AI, and 9% said they got a promotion connected to their AI skills.
That result matches what the two most careful ongoing studies on this question have found. Yale's Budget Lab tracks how AI use is changing tasks and jobs across the economy every month, and its latest reading still finds no clear connection between AI use and changes in employment or unemployment nationally. California built a similar tracker inside its state labor department, comparing unemployment claims against how exposed each occupation is to AI, and it also found no surge in AI-related layoffs so far.
There is a catch, though: a separate data set, kept by the outplacement firm Challenger, Gray and Christmas, tracks the reasons companies give when they announce layoffs, and AI has quickly become the most common excuse. Companies cited AI for about 55,000 job cuts in 2025, roughly 5% of the total. By May of 2026, AI was named as the single biggest reason for layoffs for three months running, tied to nearly 40% of all cuts announced that month, the highest share ever recorded for that reason.
Those two pictures are not as contradictory as they look. A company that blames AI for cuts sounds forward-looking to investors, while one that admits weak sales or bad planning does not. Workers, meanwhile, rarely know the exact reason they were let go, and that gap is why a worker survey and a corporate press release can describe the same wave of layoffs so differently.
The place where change is actually showing up is not existing jobs, it is the first job. Yale management professor Jeffrey Sonnenfeld and his co-authors argue that companies are not firing staff in large numbers, they are freezing hiring and squeezing more work out of the people already there, which quietly shrinks the number of entry-level openings for new graduates. Other economists push back on how much of that is AI's doing, pointing to one detailed look at graduate employment which found that almost all of the rise in unemployment among young college graduates since 2024 came from more graduates searching for work, not fewer jobs existing for them to find.
For a business owner or manager reading this, the honest takeaway is a mixed one. AI is not currently wiping out jobs across the economy, and treating every AI headline as a five-alarm fire is not backed by the data. It is already changing who gets ahead, since workers who picked up AI skills in this survey were three times more likely to report a promotion than a job loss.
The layoff label matters too: when a competitor announces AI-driven cuts, it is worth asking whether that is really what happened, or whether AI is just this year's most convenient explanation. The bigger long-term risk is not today's staff, it is tomorrow's, since fewer entry-level hires now means fewer trained managers a decade from now.