When AI Swept the Tech Ladder, Fresh Graduates Ended Up Behind Cash Registers

It feels almost like a plot twist straight out of a sitcom. In late 2022 OpenAI unleashed ChatGPT, and within months the entry‑level tech job market started to look a lot emptier. Suddenly, the career path that a computer‑science student had spent four years mapping out vanished, or at least shrank dramatically.

Two recent Census Bureau studies paint a clear picture. The first, released in April 2026, matched employer‑employee records and showed that hires for 22‑ to 24‑year‑olds plummeted in the five industries most exposed to AI. Over ten quarters after ChatGPT’s debut, early‑career employment in those sectors fell by about 12 %. Lee Tucker, one of the paper’s co‑authors, insists the drop in new hires—not layoffs of existing staff—is the main driver of the spike in unemployment among recent grads.

The hit was not random. Those AI‑exposed industries cluster around software development and information‑technology roles—exactly the jobs that computer‑science, data‑science and related majors were groomed for. A follow‑up paper, also co‑authored by Tucker, tracked graduates from the most AI‑vulnerable majors. One quarter after graduation, their odds of holding a job were five percentage points lower than peers from less‑exposed fields, and average earnings were down 13 %.

Thirteen percent might sound like a modest dip, but economists say that’s the sort of earnings hit you’d expect if a whole cohort graduated into a deep recession. The odd thing? The broader economy wasn’t in a downturn; the slump was localized to the tech‑heavy slices of the labor market.

What does a 13 % earnings loss look like on the ground? Most of these grads are still employed—they’re just not in the roles they trained for. About half of the pay gap stems from lower wages within the same industry; the other half comes from a sector switch, often into retail, restaurants or other service jobs that pay far less. In plain terms, a recent CS graduate might apply for an entry‑level coding position, find the posting gone, and end up ringing up groceries or bussing tables to pay the bills.

Even as hiring began to recover by early 2025, the rebound was on a much smaller base of openings. The gap narrows over time, but the data show it never fully disappears for the most AI‑exposed majors. Meanwhile, enrollment numbers tell a complementary story. Undergraduate CS enrollment dropped 8.4 % in spring 2026 compared with the prior year, after an earlier 3.6 % dip. A Goldman Sachs analysis flagged a 10 %+ fall in CS and programming majors for the 2025‑26 academic year—the first time economists could clearly see students reacting to AI in their major choices.

Surveys back this up. Roughly 42 % of bachelor’s‑degree students said they’d reconsidered their major because of AI, according to a Gallup‑Lumina study. And the labor market data are sobering: the Federal Reserve Bank of New York’s tracker showed a 42 % underemployment rate for recent grads in Q2 2026, with unemployment at 5.6 %—well above the national average.

Long‑term outcomes aren’t much brighter. Analyses by the Strada Institute and Burning Glass Institute reveal that more than half of graduates end up in jobs that don’t require a degree—think retail, food service, hospitality or basic office admin—within a year of leaving school, and 45 % are still there a decade later.

So the narrative is clear: AI hasn’t erased tech jobs entirely, but it has reshaped the entry point. For many new graduates, the path now winds through a grocery aisle or a fast‑food kitchen before (or instead of) a software dev desk. Whether that detour is temporary or permanent remains to be seen, but the data suggest the ripple effects of ChatGPT will be felt for years to come.