The Jobs America Is Quietly Losing to AI

In 2025, American employers blamed artificial intelligence for nearly 55,000 announced layoffs — and that number only counts the companies honest enough to say so out loud.

The Layoffs Nobody Wants to Name

Ask a company why it just cut 2,000 jobs and you will get a press release about “streamlining,” “operational efficiency,” or “evolving business needs.” What you will rarely get is the word that actually explains it. Challenger, Gray & Christmas — the outplacement firm that has tracked corporate layoffs since the 1980s — found that employers explicitly cited AI as a factor in 54,836 job cuts in 2025 alone. Since companies started copping to it in 2023, the running total has crossed 71,000. Those are just the ones willing to say the quiet part out loud.

The bigger picture is worse. Employers announced 1.2 million job cuts in 2025, up 58% from the year before and the highest total since the pandemic-scarred year of 2020. Technology led every other private sector with more than 154,000 cuts, which tells you something about who is doing the automating and who is getting automated. Amazon cut 14,000 corporate jobs in October — its largest single layoff ever — with an executive explaining the company needed to operate “more leanly” now that AI enables faster work. Microsoft cut 15,000. Neither company is struggling. Both are profitable. They are just choosing to need fewer people.

“We used to talk about jobs being shipped overseas. Now they’re just being shipped to a server rack down the hall.”

The Door Is Closing Fastest at the Bottom

If you are twenty-three and job hunting, the math is uglier than the headline numbers suggest. Entry-level job postings are down 35% since early 2023, according to labor-market analyses of listings across major hiring platforms — and more than a third of jobs still labeled “entry-level” now demand years of prior experience, which is its own quiet insult. The tasks that used to be a new hire’s on-ramp — drafting the first version, summarizing the document, writing the boilerplate code — are exactly the tasks a language model does for free, instantly, without benefits or a learning curve.

A Stanford Digital Economy Lab working paper released in late 2025 found that workers aged 22 to 25 in the occupations most exposed to generative AI saw a 16% relative decline in employment since the technology spread. Unemployment for college graduates aged 22 to 27 climbed to 5.7%, well above the 4.2% rate for the workforce overall — a reversal of the historical pattern where degrees were supposed to be the insurance policy. Underemployment among recent grads hit 42.5%, the highest share since the pandemic, meaning nearly half the graduating class is working jobs that don’t require the degree they just paid for.

Who Still Gets Hired

The damage isn’t evenly spread. Healthcare, skilled trades, and cybersecurity are still hiring new graduates at a normal clip — the jobs AI can’t yet touch a patient, wire a panel, or take legal responsibility for. But the National Association of Colleges and Employers is forecasting 2026 to be the worst hiring year for new graduates since the start of the pandemic, and the people landing offers increasingly look like they already have three internships, a demonstrated ability to work alongside AI tools, and a network that got them in the door before the posting went up. The rungs at the bottom of the ladder aren’t gone. There are just fewer of them, and they’re higher off the ground.

What makes this decline different from past waves of automation is the collar. Factory work has been automated for decades and we built (imperfect) social scaffolding around it — retraining programs, union protections, regional development funds. White-collar knowledge work was supposed to be the safe tier, the reward for doing everything right: get the degree, avoid the trades, sit at a desk. That assumption is now the thing quietly dying.

The Index tracks this because a labor market that quietly stops needing its youngest workers doesn’t announce itself with a single dramatic headline. It shows up two or three years later, in delayed marriages, in adults still living with parents, in a generation that never got the entry-level rep that used to teach them how to have a career at all. By the time the trend is undeniable, the ladder has already been pulled up.