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

Stanford researchers once reported that AI-related job sectors were seeing lower entry-level employment rates compared to other fields. However, a new study from economists at Munich's CESifo has contradicted this finding, asserting that there is no substantial evidence suggesting widespread job loss or reduced hiring for recent college graduates in both absolute and relative terms.

The research paper, titled "The Early Impacts of AI on Employment Among Recent College Graduates," was conducted by Robert Fairlie and Jane Wu. The researchers decided to concentrate on recent graduates due to the potential first indications of labor demand shifts, such as reductions in new hiring. According to the authors, as AI becomes proficient in handling standardized tasks in numerous entry-level office positions, businesses might opt to reduce hiring for simpler roles instead of laying off seasoned long-term employees.

There is some indication that job seekers graduating in 2026 might face a higher risk of AI displacement than those who graduated just a year or two earlier. The CESifo researchers note a recent surge in firms replacing numerous employee tasks with AI, as reported in a Census survey. Additionally, there has been a significant increase in AI spending per employee and ChatGPT Enterprise token usage within the past year.