A new study from Stanford and Anthropic has sparked widespread discussion by suggesting that artificial intelligence may be contributing to a drop in employment among younger workers. Researchers analyzed payroll data from ADP, covering 3.5 to 5 million U.S. employees monthly from January 2021 through September 2025, and found that workers aged 22 to 25 in roles highly exposed to AI experienced employment declines of up to 16% compared to less exposed positions. Meanwhile, older employees in the same fields saw stable or slightly rising employment, up 6–9%.
The study indicates that job losses were concentrated in roles where AI automated tasks, whereas positions where AI augmented work often saw employment growth. Notably, wages remained stable despite the shifts, and the pattern emerged around late 2022, coinciding with the launch of ChatGPT. However, the authors acknowledge that factors beyond generative AI could be influencing these trends.
Experts caution that the broader economic context is crucial for understanding these changes. The period studied also saw the U.S. Federal Reserve raise interest rates from 0% to 5.5%, ending a long era of cheap capital. Major tech companies underwent substantial layoffs: Meta cut 21,000 jobs, Amazon 27,000, Google 12,000, and Microsoft 10,000 between 2022 and 2024, totaling over 300,000 layoffs. Venture funding dropped by roughly 35% from 2021 levels, forcing startups to hire more selectively.
Even within the tech sector, the decline in hiring of younger employees mirrors broader market corrections rather than purely AI-driven displacement. Census Bureau data shows that by mid-2025 only around 5–6% of U.S. companies were actively using AI in operations. Anthropic itself notes that current AI use is just a fraction of its potential capabilities.
Historical parallels reinforce this perspective. For example, U.S. law school enrollments dropped by 28% between 2010 and 2015, and young lawyers faced extended difficulty finding jobs due to oversupply and the fallout from the 2008 financial crisis—long before AI became a factor. Similarly, pandemic-era over-hiring in tech followed by the post-ZIRP correction contributed to current workforce adjustments.
While AI will undoubtedly influence labor markets in the coming years, immediate claims that it is "killing" jobs for juniors may overstate the impact. Analysts suggest that true effects on employment will become clearer over the next three to five years as adoption expands and market conditions stabilize.
For businesses and competitors, this means the current slowdown in hiring is largely cyclical and macro-driven, not solely AI-induced. Companies should plan strategically, recognizing that the labor market may rebound once capital costs decrease and AI integration becomes more widespread.
Informational material. 18+.
