A new IZA discussion paper by Robert W. Fairlie and Jane Wu, titled The Early Impacts of AI on Employment among Recent College Graduates, says the early AI shock many expected for recent graduates is not yet visible in summer 2026 unemployment data. Using CPS microdata for June, July, and August 2026, the paper says it found no evidence of a statistically significant increase in unemployment for recent college graduates in that period.
The paper’s result is limited to unemployment in the period it studied. The supplied research says the authors frame AI’s early impact as debated and do not claim broad displacement.
In reporting on the same topic, the Associated Press described a difficult job market for new graduates, especially in software development, and said computer science graduates are emphasizing AI skills to stand out. Ars Technica summarized the working paper as finding no evidence of significant, widespread displacement or reduction in hiring.
What the paper says
According to the supplied research, the paper uses CPS microdata and examines June, July, and August 2026. In that window, the authors found no statistically significant increase in unemployment for recent college graduates. The supplied research also says there was no significant rise relative to older college graduates or young workers without a college degree.
The research bundle says the paper is cautious about interpretation and frames AI’s early impact as debated. The supplied sources do not present the paper as a final verdict on AI and graduate employment; they present it as a check on whether the expected early shock is visible yet in the unemployment data.
What the reporting adds
The Associated Press article, published on September 27, 2026, describes a difficult job market for new graduates, especially in software development. It also says computer science graduates are emphasizing AI skills to stand out. That reporting points to pressure in a specific part of the entry-level market, but the supplied sources do not say that it amounts to a measured economy-wide rise in unemployment for recent college graduates.
Ars Technica’s September 25, 2026 report places the paper against earlier claims that AI was already hurting entry-level hiring. Its summary of the working paper is narrower: no evidence of significant, widespread displacement or reduction in hiring.
Who is affected
The supplied research identifies recent college graduates as the main group studied in the paper, with comparisons to older college graduates and young workers without a college degree. The AP article specifically says the difficult market is especially pronounced in software development, and it says computer science graduates are responding by leaning harder into AI skills.
The supplied sources do not identify the full range of majors, schools, or employers affected, and they do not provide a broader workforce-composition change.
What changed, and what did not
What changed, according to the supplied research, is the tone of the conversation around entry-level AI exposure. The paper and the reporting together describe a market in which AI anxiety is high, some fields are difficult, and applicants are responding by adding AI skills. What did not change, in the paper’s data, is the measured unemployment pattern for recent college graduates in June, July, and August 2026: the authors say they found no statistically significant increase.
A finding of no statistically significant increase in unemployment does not mean there were no hiring frictions, no field-specific pressure, or no longer-run effects. The supplied sources do not identify those broader effects as established outcomes in the data reviewed here.
What to watch next
The supplied sources point to unresolved questions without answering them. The paper is limited to summer 2026 CPS microdata, so later labor-market releases could be relevant, but the sources do not specify any next benchmark date. The research also notes that the paper’s early-impact findings are about unemployment, not a broader measurement of every labor-market change.
For employers, educators, and policymakers, the current evidence suggests caution about treating entry-level hiring as an immediate AI casualty. For students, especially those in computer science and software-adjacent paths, the reporting suggests that adding AI skills is already part of how some candidates are trying to compete. The supplied sources do not say that this strategy guarantees an outcome; they only show that it is happening in a difficult job market.
Sources
- The Early Impacts of AI on Employment among Recent College Graduates — IZA@LISER Network
- As the coding boom fades, computer science grads focus on AI skills in choppy job market — Associated Press via Tulane University
- AI was supposed to hit new grads hard. So far, unemployment data says otherwise. — Ars Technica