The best current evidence links generative AI to weaker early-career outcomes for recent graduates in the fields and hiring markets most exposed to it. The clearest signal is in hiring of young workers rather than in a count of jobs lost. The evidence does not show that AI alone caused the broader slowdown in graduate job prospects, because the estimates come from different populations, states, time windows and outcome measures.
Is AI taking entry-level jobs?
The most direct evidence concerns hiring rather than layoffs. A 2026 Census Bureau working paper, built from matched employer-employee administrative tabulations, finds that hiring of workers aged 22–24 fell immediately after ChatGPT’s late-2022 launch in the industry-state cells most exposed to AI, and that the fall persisted. The authors say the hiring decline is the main cause of the large employment declines seen over the following period.
As an Amazon Associate I earn from qualifying purchases.
Three limits apply. The population is defined by age, not by degree, so this measures entry-level hiring rather than recent graduates alone. The paper is a working paper, not a settled consensus. Its historical decomposition also assigns part of the broader weakness to other forces: monetary-policy shocks through 2023 may explain up to one quarter of relative employment declines through 2025 Q2. The same analysis does not find that those shocks explain the sharp drop in hiring at the most exposed firms compared with others.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Is AI making it harder for recent college graduates to find jobs?
Two analyses that follow graduates by college major point in the same direction, but they measure different things in different places.
#1 Best Overall
Graduates by major, nationally (Census Bureau, September 2026)
A September 2026 Census Bureau working paper ranks college majors by AI exposure and compares what happened to their graduates after ChatGPT appeared. In regression-adjusted estimates for the most exposed tenth of majors, the likelihood of initial employment fell by five percentage points, and full-quarter initial earnings were 13% lower. The effects shrink with time since labor-market entry but remain substantial for the most exposed majors. Employment, earnings and job-switching patterns diverged for these majors after late 2022.
The 13% is an earnings gap, not a count of jobs. The paper attributes roughly half of the earnings decline to lower pay within the same industry sectors; the rest reflects movement into lower-wage sectors such as restaurants and retail.
Graduates in Texas (Federal Reserve Bank of Dallas, September 22, 2026)
The Dallas Fed analysis uses Texas administrative education and wage records for four-year university graduates. A 10-percentage-point higher share of automatable tasks in a graduate’s major is associated with a 1.7-percentage-point relative decline in employment within the first year. Among graduates who found work in Texas, first-year earnings in more-exposed majors fell about 5% from 2021 to 2024, relative to less-exposed majors. The event study finds that the employment-rate gap between the two groups was stable before ChatGPT’s release. Because the data cover Texas only, these are state estimates, not national ones.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Which college majors are most exposed to AI?
In the Dallas analysis, a major’s exposure comes from the occupations its graduates typically enter. Each major is scored by the AI exposure of the occupations whose job advertisements have historically asked for that major. Computer science, computer engineering and languages rank as the most exposed. Nursing, education and psychology rank among the least exposed.
Treat these rankings as averages across a whole major. A computer science graduate whose job involves little automatable work is not described by the ranking. The Census paper uses its own exposure rankings, so a major’s position in one study should not be assumed to carry over to the other. For career decisions, the more useful question is which tasks a specific job involves and whether employers are hiring for them.
Does the employer outlook point only downward?
NACE’s April 2026 employer survey summary projects 5.6% more hires for the Class of 2026, but the growth is uneven. Information, engineering services, wholesale trade, construction and miscellaneous professional services are among the industries expecting to hire more. Utilities and several manufacturing categories are among those expecting to hire less. These are employer projections rather than realized placements, and they do not show that AI creates net jobs.
What skills do employers want from new graduates now?
NACE’s summary lists teamwork, problem-solving and communication among the skills employers look for on Class of 2026 candidates’ résumés. It also says employers increasingly expect AI-related capabilities from early-career talent. The practical reading is that AI skill sits alongside judgment and collaboration rather than replacing them. A résumé or portfolio can show this with specific evidence:
Free tools Windows power users keep installed
One-click scans. No signup required.
- Problem-solving: one project where you defined the problem, weighed the options and explained why you chose one.
- Teamwork: a group deliverable, with your role stated in plain terms.
- Communication: an example where you explained a technical result to someone outside the field.
- AI use: describe the task, what the tool did and what you checked yourself, rather than a list of tool names.
None of these sources tests whether learning a particular tool improves a graduate’s chances, so no single tool should be treated as a route to a job.
How younger workers use and worry about AI
The Board of Governors of the Federal Reserve System’s 2026 report on household well-being in 2025 shows adoption and anxiety existing side by side. Across workers, 25% used generative AI at work in the prior month. Among workers aged 18–29, 20% did so, 23% worried AI would replace their job, and 19% said it would improve their career. These are self-reported survey answers about use and attitudes, not measured job losses.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Reading the labor-market numbers without mixing them up
These sources answer different questions, so their figures are not interchangeable. The table sets out how each one is built.
| Source | Population | Geography | Outcomes measured | Exposure measure | Estimate type |
|---|---|---|---|---|---|
| Census Bureau major-outcomes working paper (September 2026) | Recent graduates, grouped by major | Not stated | Initial employment likelihood; full-quarter initial earnings; job switching | Major-level AI exposure rankings | Regression-adjusted comparison of the most exposed decile |
| Federal Reserve Bank of Dallas (September 22, 2026) | Four-year university graduates, by major | Texas | First-year employment; first-year earnings | Automatable tasks in occupations whose job advertisements historically asked for each major | Event study around ChatGPT’s release, with association estimates |
| Census Bureau early-career hiring working paper (2026) | Workers aged 22–24 | Industry-state cells | Hires; employment | Employer and industry AI exposure | Comparison across exposed and less-exposed cells; historical decomposition |
| New York Fed recent-graduate series | Bachelor’s degree holders | National | Unemployment; underemployment | Degree requirements from O*NET data | Descriptive series, generally updated quarterly; outcomes by major updated annually |
| Bureau of Labor Statistics October 2024 snapshot (released April 22, 2025) | Recent graduates aged 20–29 who earned degrees January–October 2024 | Not stated | Employment; unemployment; enrollment in school | Not applicable | Single snapshot for October 2024 |
| Board of Governors of the Federal Reserve System household well-being report (2026) | Workers, including those aged 18–29 | Not stated | AI use at work; worry about replacement; expected career benefit | Not applicable | Self-reported survey |
| NACE employer survey summary (April 2026) | Employers hiring the Class of 2026 | Not stated | Projected hiring change; skills and AI expectations | Not applicable | Employer projections |
BLS October 2024 snapshot
The Bureau of Labor Statistics, in a release dated April 22, 2025, reports October 2024 outcomes for recent graduates aged 20–29 who earned degrees during January–October 2024. Among recent bachelor’s recipients, 69.6% were employed, 15.3% were unemployed and 25.2% were enrolled in school. Continued enrollment matters when describing post-graduate paths, because the employed and unemployed shares alone leave out graduates who are still in school.
Recommended Free Tools
New York Fed recent-graduate series
The New York Fed’s national series covers bachelor’s degree holders from 1990 to the present and is generally updated quarterly, with outcomes by major updated annually. It defines underemployment as a recent college graduate working in a job that typically does not require a college degree, using U.S. Department of Labor O*NET Education and Training Questionnaire data to judge degree requirements. Underemployment is a different measure from unemployment, and the series is not an AI measure. It is useful background on the wider market.
Is AI a silent threat, or a measurable one?
The title’s word “silent” fits the mechanism better than the measurements. The effects appear in hiring decisions and starting pay for particular groups, and the aggregate series above do not isolate them. Two open questions remain:
Quick Recap
- Causal share. The studies estimate associations and event-study patterns. None produces a count of jobs AI has eliminated.
- Duration. The estimates cover the period since ChatGPT’s late-2022 launch. Whether these gaps persist for later graduating classes is not established by these sources.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
