A revised Stanford University study finds that artificial intelligence is significantly reducing job opportunities for young workers in occupations most exposed to AI, even as older employees remain largely unaffected. The updated paper, an August 2026 revision of last year's "Canaries in the Coal Mine?" research, shows the gap between AI-exposed and AI-resistant fields has widened, suggesting the disruption is not a one-off blip but a deepening trend with implications for graduates entering the workforce.

Using ADP payroll data, researchers found that employment for workers aged 22 to 25 in the most AI-exposed occupations is now 19% lower than in less-exposed fields, up from a 13% gap a year earlier. Since 2022, employment in the most AI-impacted jobs has fallen roughly 11% for this age group, while it has risen about 10% in the least-affected jobs, with the effect driven mainly by reduced hiring rather than layoffs or pay cuts. The researchers distinguish between "automative" AI uses, which replace human tasks (hitting roles such as accountants and receptionists hardest), and "augmentative" uses that support workers (as seen in roles like chief executives and nurses), with automation-heavy jobs showing the weakest employment outcomes for entry-level staff.