Computer Science Enrollment Is Crashing — and What It Means for the Next Generation of Developers

Last month, the National Student Clearinghouse dropped preliminary fall enrollment data, and the number was hard to ignore: undergraduate computer and information science programs fell more than 12 percent this semester. Graduate programs dropped more than 10 percent. Those aren’t blips — they’re continuations of a decline that started last year and has now erased nearly two decades of growth in CS education.

I’ve spent most of my career inside government tech, managing teams that build and maintain actual systems. Right now I’m also pursuing a Master’s in Information Technology. So when I see enrollment numbers like this, I don’t just see a trend piece. I see a question about whether the pipeline that feeds our field is about to go through a painful correction — and whether anyone is preparing for what comes after.

What the Numbers Actually Say

The Clearinghouse data covers about 7.92 million enrollment records from roughly 46 percent of U.S. institutions. Four-year undergraduate CS enrollment fell 12.4 percent. Two-year programs dropped 12.3 percent. Graduate enrollment declined 10.4 percent. Over the past two years combined, undergraduate CS is down 18.4 percent, and graduate CS is down 27.1 percent.

Those figures are confirmed by independent sources. Stanford economist Jacob Light, who analyzed more than 50 million course sections across over 1,000 schools, found that CS major share declined in 2025-26 for the first time in roughly twenty years. The Bureau of Labor Statistics weighed in too: its August report identified computer and mathematics occupations as the jobs most exposed to AI displacement.

The decline isn’t uniquely American. The UK saw computing-degree enrollment among 18-year-olds fall 9 percent in 2025. Australia’s figure is even steeper — a 21 percent drop heading into 2026, per Australasian Conference of Tertiary Admissions Centres data.

Why Students Are Walking Away

The short answer is that the return on investment suddenly looks uncertain.

For years, a CS degree was treated as a golden ticket. Tech companies were hiring aggressively, entry-level coding roles were plentiful, and even mid-career switchers could land solid offers after a bootcamp or two. That market has shifted. Over 120,000 tech jobs have been cut in 2026 alone, per Layoffs.fyi, and AI was cited in more than 54,000 of those layoff announcements. Companies like Meta and Google slashed headcount while simultaneously increasing AI infrastructure spending — a signal that the work is being consolidated, not eliminated.

Here’s what that looks like on the ground: entry-level coding roles, the traditional first rung for CS graduates, are being automated or absorbed by AI-assisted senior engineers. Existing teams can ship more code with fewer juniors. A company that once hired five associates to handle tickets, bug fixes, and routine feature work might now keep two seniors and lean on coding assistants for the rest.

Students see this. They watch tech companies announce layoffs alongside record AI investment. They read headlines about AI passing coding benchmarks. And they start asking whether a four-year CS degree will still pay off by the time they graduate. When the answer feels ambiguous, the safer bet — at least in the short term — looks like nursing, engineering, or something with a clearer hiring pipeline.

Community college enrollment tells an interesting counter-story. Two-year CS programs actually rose 1.2 percent this fall. That’s a reminder that not everyone is abandoning the field — some are just choosing cheaper, faster entry points rather than a full bachelor’s commitment.

The AI Angle Is Real, but It’s Not the Whole Story

It’s tempting to pin the whole decline on ChatGPT. The timing lines up: enrollment peaked around 2023-24, and generative AI exploded into public awareness in late 2022. Stanford’s Light cautioned against treating the correlation as causation, noting that the data can’t isolate whether students are reacting to AI specifically, weaker tech hiring, or broader shifts in how they evaluate degrees.

I think he’s right to be careful, but the pattern is hard to dismiss entirely. China’s AI bubble burst and the broader AI investment boom are creating an atmosphere where the tech labor market genuinely feels unstable. When the largest employers in the field are cutting jobs and redirecting money toward automation, students notice — even if the causal chain is messy.

There’s also a migration happening within the field rather than away from it. Some students who might have majored in CS five years ago are now targeting data science, AI/ML, or analytics programs instead. The Washington Post’s analysis of Clearinghouse data found that some of the “missing” CS students are being funneled into adjacent disciplines. That’s not necessarily a tragedy — it’s a sign that students are recalibrating toward where they see actual demand.

What This Means for the Industry

I’ve interviewed candidates for government ICT roles. I’ve seen the gap between what universities produce and what teams actually need. The CS enrollment boom never fully solved that mismatch — it just flooded the market with graduates who could write syntax but couldn’t debug a production incident or talk to a non-technical stakeholder.

A correction in enrollment isn’t inherently bad if it forces programs to get sharper. The danger is that the correction is too fast and too broad, leaving gaps in places that matter. We still need people who understand networking, security, database design, and systems architecture — not just prompt engineering or fine-tuning. If students abandon CS entirely because the entry-level market looks grim, we risk losing a generation of infrastructure builders.

CIO magazine’s recent survey of 1,200+ tech leaders across 53 countries drove this home: 45 percent expect lower demand for entry-level CS grads five years from now, and the reason cited across every demographic was AI automating routine technical work. That’s not a temporary dip — it’s a structural replacement.

What I’d Tell a Student Right Now

If a high school senior or career changer asked me whether to study CS, I wouldn’t tell them to run for the hills. But I would tell them to study strategically.

First, treat the degree as a foundation, not a guarantee. The pieces that AI is worst at — debugging complex systems, understanding context across large codebases, communicating with users, making architectural tradeoffs under constraints — are exactly the skills that separate mid from senior engineers. Those skills still need human development.

Second, get comfortable with AI tools early. Running a local AI assistant or working on projects that integrate AI isn’t just resume fodder — it’s becoming table stakes. Students who graduate knowing how to pair AI capabilities with real engineering judgment will have a better story to tell than those who avoided the tools entirely.

Third, build things that matter beyond assignments. The graduates I’ve seen succeed in government tech weren’t the ones with the highest GPAs — they were the ones who maintained a community tool, contributed to an open-source project, or built something their department actually used. Credibility comes from shipped work, not transcript entries.

The Bottom Line

Computer science isn’t dying. But the era where anyone could stumble into a coding job with a generic degree is ending — and that’s probably healthy. The programs that survive this enrollment crash will be the ones that stop selling a degree as a ticket and start selling it as training for a craft.

For students, the message is clear: the market is repricing what it values. Don’t panic. Adapt. Focus on the skills that AI amplifies rather than replaces. And if you’re serious about this field, the crash might actually be the best filter we’ve had in years — it separates the people who want to build things from the people who just wanted a safe paycheck.

The pipeline is shrinking. That means the people who stay will matter more.