Artificial intelligence is indeed creating a tighter labor market for early-career jobseekers, according to new research from three Stanford University economists.
Since generative AI has become widely used, the working paper found, young workers (ages 22–25) in the most AI-exposed occupations, such as software developers and customer service reps, have experienced a 13% relative decline in employment. In contrast, jobs as health aides have been growing faster for young workers than for older ones.
Likewise, declines in entry-level jobs are concentrated in occupations where AI is more likely to automate human labor rather than augment it.
“It appears what younger workers know overlaps with what LLMs can replace,” Erik Brynjolfsson, director of the Stanford Digital Economy Lab and one of the study’s co-authors, said in an interview with Derek Thompson, a writer at The Atlantic who was early to spot this trend.
The Stanford study tapped data from ADP, a large payroll processing firm. Payroll data is a particularly good way to track complex, fast-moving changes to jobs, one labor market researcher told me. The findings also hold up when controlling for firm-level interest rate sensitivity, remote work, the pandemic’s impacts on education, and the slowdown in tech hiring.
“Economy-wide employment continues to grow, but employment growth for young workers has been stagnant,” the paper concludes.
The new research adds firm evidence about AI’s employment effects to anecdotes from large companies, dramatic predictions from tech leaders, and a smattering of credible takes about diminishing entry-level jobs and tightening career prospects for recent college graduates.
“You can’t know for sure, but it does look like AI is behind these big changes,” Brynjolfsson told Axios AI+.
The research findings, of course, won’t stop people from arguing about how AI will reshape jobs, or what policymakers, educators, and employers should do about it. For his part, Brynjolfsson says addressing AI’s impacts might require rethinking how young workers learn on the job.
“I think that we’ll have to more explicitly train people, as opposed to just hoping that they will figure these things out on their own,” he told The Wall Street Journal.
Related Research: A recent report from researchers at the Burning Glass Institute and the Project on Workforce from Harvard University looked at how AI could remap learning curves between entry-level and experienced workers while shaking up how expertise is built. The research concludes that the technology will trigger a sweeping reevaluation of professional development and human potential:
“In many fields, workers will need to find new ways to get on the career ladder while employers will need to find new ways to develop and retain talent as conventional pipelines are disrupted.”
Also this week, two researchers at the Federal Reserve Bank of St. Louis weighed in on whether AI is contributing to rising unemployment. By using measures of both theoretical exposure to AI and actual adoption of the tech, they found a “striking correlation between AI’s prevalence and unemployment increases since 2022.”
Computer and mathematical occupations saw some of the steepest unemployment rises while blue-collar jobs and personal services roles experienced relatively smaller increases. Workers in tech roles are finding that AI tools can accelerate certain tasks, they note, potentially at the cost of overall employment demand.
“Our results suggest we may be witnessing the early stages of AI-driven job displacement,” the researchers write. “Unlike previous technological revolutions that primarily affected manufacturing or routine clerical work, generative AI can target cognitive tasks performed by knowledge workers—traditionally among the most secure employment categories.”