Stanford Study Shows AI Hits Young Workers Hardest

A new study from Stanford's Digital Economy Lab looks at how generative AI is changing the job market, especially for young people just starting their careers. Using payroll data from ADP covering millions of American workers, the study doesn't find massive job losses across the country, but it does show fewer entry-level jobs.

In fields like software development and customer service, where AI has a big impact, employment for people aged 22 to 25 is about 19% lower compared to other areas. In 2025, that gap was only 13%. More experienced workers in the same jobs are doing fine—their employment is steady or even growing.

The researchers say entry-level jobs rely a lot on "codified knowledge," which is formal knowledge that can be taught through books or written rules. AI can easily do this kind of work. Jobs that need "tacit knowledge," learned through practice and mentoring, are safer—AI helps experienced workers but can't replace their judgment. The study also finds that having a college degree helps protect against AI's impact.