Research Challenges Sam Altman's Job-Creation Claim as AI Squeezes Opportunities for Young Workers

Stanford Digital Economy Lab researchers used administrative payroll records covering millions of US workers at ADP to examine labor-market shifts since generative AI became widely used. The study found no evidence of economy-wide job displacement. Instead, weakness was concentrated among early-career employees in occupations where AI can automate tasks based on codified knowledge. Experienced workers, whose roles rely more on tacit knowledge, and workers in jobs where AI complements human labor generally fared better.

The revised study, released on Aug. 12, 2026, extended the data through June. Employment among 22- to 25-year-olds in highly AI-exposed occupations was 19% below the level implied by the trend among less-exposed peers, widening from a 15% shortfall in July 2025. From November 2022 to June 2026, employment in the two most-exposed groups fell about 11%, while the three least-exposed groups gained roughly 10%. The divergence mainly reflected weaker hiring, not higher separations, and the authors cautioned that the findings are descriptive rather than causal.