The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
For some new graduates, the first job is harder to get, and the pressure falls hardest on fields whose entry-level tasks are easiest for AI to perform. Recent U.S. studies link greater AI exposure to weaker early employment, hiring, and pay. They do not show that AI alone explains the broader difficulty of the graduate market, and they do not settle how AI will change total employment over the long run.
Where AI shows up first: the hiring door, not the layoff line
AI’s clearest effect on recent graduates appears at the point of entry. A Census Bureau working paper on early careers attributes most of the drop in employment among workers aged 22 to 24 in the most exposed industry-state group to reduced hiring, not to layoffs of people already on payroll. A company that stops backfilling a junior analyst role, or that routes first-draft writing and routine coding to software, changes its intake without announcing a single job cut.
Exposure is a property of tasks, not of degrees or job titles. A major or industry is more exposed to the extent that the work a beginner typically does, such as drafting, summarizing, routine coding, or data preparation, can be done with AI. The same major can therefore carry different risk at different employers. Two marketing graduates may face very different first-year workloads: one producing content at volume, the other running campaign analysis that requires client judgment.
Recommended Free Tools
Three U.S. studies, three different questions
The three studies below use different populations, periods, and exposure definitions. Their numbers answer different questions and should not be added together or read as one effect size.
| Study | Who and when | How exposure is defined | Headline estimate | Main qualification |
|---|---|---|---|---|
| Census Bureau working paper CES-WP-26-56 (September 2026) | Recent college graduates by major, measured at initial labor-market entry | Decile of majors ranked by AI exposure | Most-exposed decile: five-percentage-point decline in likelihood of initial employment; 13% decline in full-quarter initial earnings (regression-adjusted) | Estimates for exposure-defined groups, not a forecast for each graduate |
| Census Bureau working paper CES-WP-26-27 | Workers aged 22 to 24, over the ten quarters following ChatGPT’s introduction | Industry-state quintile, most exposed group | Adjusted employment down 12% in the most exposed quintile; reduced hiring was the primary contributor | Hiring rates largely recovered by early 2025, on a smaller employment base; the paper also flags possible earlier trend shifts around COVID and discusses remote work and educational attainment as possible explanations |
| Federal Reserve Bank of Dallas analysis | Four-year graduates of Texas universities | Share of automatable tasks associated with each major, using the Dallas Fed’s own Texas measure | A 10-percentage-point higher automatable share was associated with a 1.7-percentage-point relative employment decline within a year; first-year earnings of employed graduates in more-exposed majors fell about 5% from 2021 to 2024, relative to less-exposed majors | Texas only; an association between exposure and outcomes, not a national effect size |
Reading the major-level findings
The Census major-level estimates speak most directly to the entry transition. The paper reports that effects attenuate as graduates move further from entry, while remaining substantial for the most exposed majors. That suggests the penalty narrows with experience, but it does not show that the early gap disappears in the most exposed fields.
Reading the hiring findings
The 12% figure describes a decline in employment level, not a permanent collapse in hiring. Hiring rates largely recovered by early 2025, but from a smaller employment base, so fewer young workers hold these jobs than before. Because the paper also raises trend shifts that predate ChatGPT, the 12% is best read as an estimate tied to AI exposure rather than a measure of how much ChatGPT alone cost young workers.
Reading the Texas findings
The Dallas Fed analysis is the only one here that ties outcomes to one state’s graduates, and it uses its own exposure measure. Computer science, computer engineering, and languages ranked among the more-exposed majors; nursing, education, and psychology among the less-exposed. Those rankings depend on the Dallas Fed’s task measure, and a different exposure index could order majors differently.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallThe wider graduate market: observed conditions and employer plans
The New York Fed’s national series on recent graduates, updated quarterly, shows conditions that remain difficult in 2026 Q2. The series is available at the New York Fed’s college labor market page. Separately, NACE’s April 2026 Spring Update reports that employers project 5.6% more hiring for the Class of 2026, with results described as uneven across industries and employers. One set of figures measures what has happened; the other measures what employers expect.
| Measure | Type | Period | Value | What it does not show |
|---|---|---|---|---|
| Recent-graduate unemployment, New York Fed national series | Observed | 2026 Q2 | About 5.6% | Causes; it does not isolate AI from other factors |
| Recent-graduate underemployment, same series | Observed | 2026 Q2 | 42% | Unemployment. It counts degree-holders in jobs that typically do not require a bachelor’s degree |
| Class of 2026 hiring projection, NACE Spring Update (April 2026) | Employer expectation | Class of 2026 | 5.6% projected increase in hiring | Realized hiring; the projection varies by industry and employer |
The unemployment rate and the NACE projection both come to 5.6%, but they are unrelated measures. A market can be hard for graduates while employers still plan to hire more of the Class of 2026, and these figures show both at once.
What employers say they are looking for
NACE reports that employers want evidence of teamwork, problem-solving, and communication on Class of 2026 resumes. A Federal Reserve Board FEDS Note from March 2026 describes demand for AI skills as spreading beyond computer and mathematical occupations. Taken together, that points to three practical implications:
- AI literacy works as a readiness signal, useful in roles where these tools are routine.
- It is not a guarantee of employment, and it does not protect a specific task from automation.
- Human skills such as teamwork and communication are persuasive when shown through examples rather than claimed in a summary line.
Does more education help?
The Dallas Fed analysis reports that more-exposed Texas graduates were more likely to return to graduate study. That describes behavior, not a payoff. Its evidence suggests that formal upskilling within an exposed field offers limited returns unless the new expertise complements what AI does. A program is more likely to pay off when it adds domain judgment that AI output still has to be checked against, and less likely when it mainly teaches a tool.
Check your own exposure before choosing a path
These questions follow the task-based logic of the studies. They are a diagnostic for thinking about your own position, not a validated score.
- What share of a typical first-year week would be drafting, summarizing, routine coding, data entry, or similar production work?
- Has your target employer posted junior roles recently? Check current postings rather than general reputation.
- Can you show a finished piece of work that a hiring manager could review? If not, that gap may matter more than the name of your major.
- Does the job require you to judge, verify, or take responsibility for AI-produced output?
Preparation that improves your position
None of the steps below guarantees a job or protects against automation of specific tasks. They make it easier to show value in an entry-level market that is selective.
- Build work samples. Produce two or three pieces that show your reasoning, such as a data analysis with a written conclusion, a product teardown, or a working prototype with documentation.
- Get work in front of real stakeholders. Internships, co-ops, freelance projects, and volunteer work with deadlines and feedback give you evidence an employer can verify.
- Document AI-assisted work. For each project, note what the tool produced, what you checked or corrected, and why you made the final call. That record demonstrates the judgment employers describe.
- Put teamwork, problem-solving, and communication on the page. Use concrete examples, such as a disagreement you resolved in a group project or a decision you explained to a non-technical audience, rather than listing the words.
What remains unsettled
No agreed estimate exists for AI’s long-run net effect on graduate jobs or total employment. The Federal Reserve Board’s note says work on aggregate employment effects is still early and mixed. The studies above cover specific populations over specific periods, so they cannot yet show how the entry-level gap will look after several more years of AI adoption.
Three signals would sharpen the picture: whether the New York Fed’s recent-graduate series improves from its 2026 Q2 level; whether NACE’s Class of 2026 projection is borne out by realized hiring; and whether later Census Bureau or Dallas Fed updates show the early-career gap narrowing or widening.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Quick Recap
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.
