AI was supposed to hit new grads hard. So far, unemployment data says otherwise.

Key Points:

• A new CESifo working paper challenges previous findings by a Stanford study, reporting no significant reduction in hiring or increased unemployment among recent college graduates due to AI impacts as of summer 2026.
• Researchers Robert Fairlie and Jane Wu analyzed US Census Current Population Survey data, focusing on recent graduates aged 22 to 25, and found the 2026 unemployment rate (7.3%) within historical norms, even when accounting for those seeking jobs but not officially unemployed.
• Statistical comparisons with non-college graduates and older college graduates showed no significant differences in employment trends related to AI exposure from 2022 to 2026, suggesting AI has not yet displaced new graduates on a broad scale.
• The discrepancy with the Stanford study may stem from differences in data sources and metrics, as Stanford used payroll data from ADP focusing on job supply, while CESifo examined unemployment rates reflecting job demand.
• The CESifo researchers caution that while 2026 data shows limited AI impact on employment, increased AI adoption in workplaces could affect future graduating classes, necessitating ongoing monitoring of labor market trends.