Stanford warns AI is reshaping the job market — and Gen Z is taking the hit

The first tremors of the AI revolution are already shaking the U.S. labor market, according to Stanford University – and young workers are feeling it most.

A new study found that employment among 22- to 25-year-olds in the most AI-exposed jobs, such as software engineering and customer service, has dropped 13% since late 2022. The research, based on payroll data from millions of workers, suggests that generative AI is displacing entry-level employees even as older workers in the same roles hold steady or see gains.

The findings add hard data to a growing sense of anxiety among Gen Z. Goldman Sachs recently calculated that the value of college degrees is quickly diminishing, while Bank of America Global research has said that the unemployment rate for recent graduates has started to exceed the overall rate for the first time in years.

In occupations less affected by AI, hiring patterns for younger and older workers have looked similar to one another. But in roles more directly exposed to the technology, employment among people aged 22 to 25 has slipped about 6% since late 2022. Older workers in the same fields, by contrast, recorded gains of 6% to 9%.

Researchers said it amounted to “early and large-scale evidence” of a labor-market shift, and argued that “experience and tacit knowledge are becoming crucial buffers against displacement as AI systems excel at replacing book learning over job-specific, hard-to-codify skills.”

Job cuts are hitting knowledge workers especially hard. There have been large waves of layoffs at companies such as Microsoft, Duolingo, and Walmart. At the same time, many companies have spent big on the technology. At Duolingo, CEO Luis von Ahn announced an initiative in April to make the language-learning app “AI-first.”

Not all AI adoption is hurting jobs, however, the Stanford researchers continued. They said declines were concentrated in occupations where AI “substitutes” for human labor, such as in customer service or some basic coding roles. 

By contrast, “augmentative” uses – such as AI that flags anomalies in medical scans or drafts text for marketers to refine – tend to boost productivity without cutting entry-level jobs, the report said. “These findings are consistent with automative uses of AI substituting for labor while augmentative uses do not,” it said.

It comes after separate polling by Ipsos and Reuters suggested that as many as 71% Americans are deeply concerned that AI could put swathes of the country out of work. And a recent study by researchers at Oxford University and Deloitte found that about 35% of current jobs are at high risk of automation over the following 20 years.

“Our findings suggest that generative AI is not yet having uniform effects across the labor market, but is instead reshaping employment opportunities in ways that disproportionately disadvantage young workers,” the Stanford researchers added. “We view these findings as the beginning of the AI revolution in the labor market.”