New research from CESifo, a Munich-based economics institute, challenges the growing narrative that AI is already hollowing out entry-level jobs for college graduates. The working paper, "The Early Impacts of AI on Employment Among Recent College Graduates," found no evidence of significant hiring displacement for this group through summer 2026. This matters for HR leaders who are building workforce plans around assumptions of AI-driven restructuring.
Researchers Robert Fairlie and Jane Wu focused on recent graduates because, as they argue, "changes in labor demand may first appear through reductions in hiring." The logic is straightforward. Firms might stop hiring for roles with standardized tasks that AI can handle, rather than laying off experienced employees. Yet the data tells a different story.
The numbers behind the conclusion
The team analyzed microdata from the US Census Current Population Survey, tracking unemployment among Bachelor's degree recipients aged 22 to 25 who are not pursuing higher degrees. They looked at year-over-year and seasonal trends going back to 2022, the year employment returned to pre-pandemic levels and ChatGPT launched.
The summer 2026 unemployment rate for these graduates hit 7.3 percent. That figure sits squarely within the range seen in previous years-6.3 percent in 2022 and 7.8 percent in 2024. The numbers remained unremarkable even when the researchers included graduates who told the survey they wanted a job but were not actively looking.
The researchers then ran statistical comparisons against non-college graduates of the same age and older college graduates aged 30 to 49. They also broke down employment by potential "AI exposure" using a 2023 study of which roles AI systems are best equipped to handle. Across nearly all comparisons, any trend differences between groups from 2022 to 2026 were not statistically significant. The data, they wrote, "tell a consistent story in which unemployment among recent college graduates in summer 2026 was not unusually high relative to earlier summers."
Contradicting the Stanford study
These findings run counter to a Stanford University study reported last month, which found entry-level employment in AI-impacted occupations lagging behind other fields. The divergence likely comes down to methodology. Stanford's study used payroll data from ADP, which covers a large slice of the economy but may miss elements captured by the broader Census survey. The ADP data also measures total job supply in various fields, while the unemployment rate studied here reflects aggregate demand for those jobs. Demand can shift even as supply contracts.
Why 2026 might be different-and what comes next
There are reasons to suspect 2026's graduates could face more risk than those from a year or two earlier. The CESifo researchers point to a sharp increase in firms "replacing a large number of employee tasks with AI" in a Census survey, along with broad increases in AI spending per employee and ChatGPT Enterprise token use over the past 12 months.
Some prominent voices expect this to be the inflection point. Venture capitalist Marc Andreessen said earlier this year that "AI literally until December [2025] was not actually good enough to do any of the jobs that they're actually cutting." BlackRock CEO Larry Fink said in March that "the speed at which AI is changing" led him to worry that "when this year's college graduates enter the workforce, we could see the highest unemployment rate among them in years-even without a recession."
The CESifo researchers view their analysis as a "useful first test" of how accelerating AI usage is-or is not-affecting the job market. They caution that current trends do not predict future performance. If workplace AI intensity keeps increasing, the graduating classes of 2027 and later might face different conditions. Additional years of data will be needed to detect any effects as AI use deepens.
Why this matters for HR and management
For HR leaders, the CESifo paper provides a data point against panic-driven restructuring of campus recruiting programs. The absence of widespread displacement through mid-2026 suggests that claims of an immediate AI-driven hiring collapse are not supported by broad labor market data. This buys time to develop deliberate workforce strategies. Professionals looking to build that capability can explore structured learning like AI VP HR Courses or AI HR Leadership Courses to align talent planning with what the data actually shows, rather than what headlines claim.
Your membership also unlocks:
