Saturday, September 26, 2026
Tech Beat

AI Has Not Raised 2026 Graduate Unemployment, CESifo Finds

CESifo finds no unusual summer 2026 unemployment among recent US college graduates, despite rising AI use and warnings of entry-level job displacement.

Listen to this briefingAudio briefing

Summary

A Munich CESifo working paper, “The Early Impacts of AI on Employment Among Recent College Graduates,” by economists Robert Fairlie and Jane Wu finds no significant, widespread AI displacement or hiring reduction, in absolute or relative terms, among recent US graduates. Using US Census Current Population Survey microdata, it defines them as bachelor’s degree holders aged 22 to 25 who are not pursuing higher degrees, then analyzes year over year and seasonal unemployment from 2022, when employment returned to prepandemic levels and ChatGPT launched, through summer 2026. Unemployment was 7.3%, within the 6.3% in 2022 to 7.8% in 2024 range. Including people who wanted jobs but were not actively looking changed little.

Tests against noncollege peers aged 22 to 25 and college graduates aged 30 to 49, plus job exposure classifications from a 2023 AI capabilities study, found almost no statistically significant trend differences. That conflicts with Stanford University research finding weaker entry-level employment in AI-impacted occupations. Stanford used ADP payroll data to measure jobs across fields, while CESifo’s broader Census measure combines job supply with worker demand, which can shift even as AI contracts some fields.

Risks are rising: a recent Census survey showed a sharp increase in firms replacing many employee tasks with AI, while AI spending per worker and ChatGPT Enterprise token use increased during the previous 12 months. Marc Andreessen said AI was not capable enough for jobs being cut until December 2025. BlackRock CEO Larry Fink warned in March 2026 that graduates could face their highest unemployment in years without a recession. CESifo calls summer 2026 a first test, not a forecast: deeper workplace adoption could hurt classes of 2027 and later, requiring more years of data.

Positives

  • Summer 2026 unemployment among recent college graduates was 7.3%, within the 6.3% in 2022 to 7.8% in 2024 range.
  • Almost all comparisons with noncollege peers and older graduates showed no statistically significant employment trend differences from 2022 to 2026.
  • Including graduates who wanted jobs but were not actively searching produced similarly ordinary unemployment results.
  • CESifo found no significant, widespread displacement or hiring reduction among recent graduates in either absolute or relative terms.

Risks & concerns

  • A recent Census survey showed a sharp increase in firms replacing large numbers of employee tasks with AI.
  • AI spending per employee and ChatGPT Enterprise token use rose broadly during the previous 12 months.
  • Stanford University’s ADP payroll analysis found entry-level employment lagging in occupations considered highly exposed to AI.
  • Larry Fink warned in March 2026 that graduate unemployment could reach its highest level in years even without a recession.
  • CESifo cautioned that deeper AI adoption could affect the graduating classes of 2027 and later more severely.
Primary sourceAI - Ars Technicahttps://arstechnica.com/ai/2026/09/ai-was-supposed-to-hit-new-grads-hard-so-far-unemployment-data-says-otherwise/
Read full article
Editorial note: Tech Beat summarizes and analyzes third-party reporting. The source link is the authoritative article. This page does not reproduce the full source text.

More From The Wire

AI InfrastructureSep 25

Anthropic Commits $11.6 Billion to Akamai Cloud in Seven Year Deal

Artificial IntelligenceSep 25

Proaction Says Codex Drives 60% Sales Jump, Saves 75+ Hours

Startups and Venture CapitalSep 25

Mark Wahlberg to Detail Startup Investing Strategy at TechCrunch Disrupt 2026