1 in 6 Students Changed Majors Over AI. The One Called "Safe" Fell Hardest. | Edition 319

Students moving back onto US campuses this month are carrying a question the class of 2015 never packed: will my major still make sense by graduation? That is no longer just a mood. It is showing up in hard enrollment data, and it is redrawing the map of what Americans study.

Two large surveys put numbers on the anxiety this year. In a Gallup study for the Lumina Foundation, 16% of enrolled US college students — about 1 in 6 — said they had already changed their major or field of study because of the effect AI may have on jobs. Nearly half (47%) said they had given the switch at least "a fair amount" of thought. And in a June–July 2026 survey of 2,000 US students commissioned by Superhuman (the company formerly known as Grammarly), 69% said they worry AI will make it harder to find a job, and 22% said job-market fears had already pushed them to change majors or concentrations.

The "safe" major cracked first

For roughly 15 years, "learn to code" was the closest thing American higher education had to guaranteed advice. The Washington Post, working from federal degree data, notes computer science has been among the fastest-growing majors in the country since the Great Recession — four-year CS degrees granted roughly quintupled between 2008 and 2024.

That run just hit a wall. National Student Clearinghouse data shows undergraduate enrollment in computer and information sciences at four-year colleges fell 8.1% in fall 2025 — from about 659,700 students to about 606,100 — the steepest one-year decline of any field of study since at least 2020. Graduate enrollment in the field dropped 14%. The traditional computer science major specifically fell 11.2%. In a single year, computer science slipped from the fourth-largest undergraduate major to sixth.

Stanford economist Jacob Light, analyzing 50 million course sections across 1,019 US colleges, found CS's share of enrollment declining for the first time in about two decades. One honest caveat he flags: the timing lines up with ChatGPT, but the data does not prove AI caused the retreat — CS has always been a boom-and-bust major. What is not in dispute is that students are reading the same headlines you are: recent grads in computer science and computer engineering have posted some of the highest unemployment rates of any majors.

Where the students went instead

They did not flee technology — they moved sideways. Engineering enrollment rose 7.3% in the same period, with mechanical up 11.4% and electrical up 13.8%; department chairs told the Washington Post students see engineering as a more flexible bet in an AI economy, with paths into robotics, aerospace, drones, and energy. Data science and cybersecurity programs are flat or growing while traditional CS shrinks.

The other destination is AI itself. Federal data analyzed by the Computing Research Association counted just 4 US schools offering an AI degree in 2020. Researchers at Northeastern University now count 79 institutions with an AI major — and more than 100 with a minor or concentration, per the Boston Globe. MIT's "Artificial Intelligence and Decision Making" major, launched in 2022, is already the school's second-most-popular program.

And then there is the creative end of the spectrum: Arizona State University's Cronkite School now offers a Bachelor of Arts in Content Creation that, in the program's own words, prepares students to become influencers — with a capstone requiring a real account, real content, and demonstrable audience growth. Critics call it unworthy of academic attention; defenders point out the creator economy is a real labor market and the skills (audience research, production, distribution) transfer.

The part the panic gets wrong

Here is the nuance most coverage skips. In a June 2026 analysis, economists at the Federal Reserve Bank of St. Louis wrote that "AI adds an additional headwind at the point of labor market entry, particularly for recent college graduates, but its effects remain smaller than those of the broader decline in job openings." Their numbers: between April 2023 and late 2025, unemployment for 18-to-24-year-olds rose 2.9 percentage points because of overall job scarcity — and 1.1 points because employers shifted demand toward AI skills. AI is real; the hiring freeze is bigger.

The New York Fed's tally makes the squeeze concrete: recent-grad unemployment stood at 5.6% in December 2025, against roughly 4.2% for workers overall, and 42.5% of employed recent grads were working jobs that do not typically require a degree.

So what should anyone actually study? OpenAI's VP of education, Leah Belsky, told Axios employers will increasingly hire for transferable skills — critical thinking, AI literacy, leadership, collaboration — rather than a specific "AI-proof" major. That is quietly becoming the consensus answer: the major matters less than whether you can use these tools with judgment and defend your work.

What this means if you are not 19

If you are a parent fielding a panicked call from campus, or a professional wondering if your own degree still holds, the same three findings apply. One: the anxiety is rational — AI is a measurable headwind for new entrants. Two: it is smaller than the headlines suggest, and dwarfed by an ordinary slow hiring market that will eventually turn. Three: the people navigating it best are not picking a magic major — they are stacking a durable skill (a field they actually like) with fluency in the tools everyone will be expected to use. Nobody future-proofed a career by guessing; plenty have by staying adaptable.