The Great CS Exodus: Enrollment Is Down, AI Is Up

Computer science enrollment at University of California campuses fell 6% in a single year — after a 3% drop the year before — even as overall college enrollment rose 2% (TechCrunch, Feb 2026; SF Chronicle data). The students are not leaving technology at all. They are leaving the plain CS major for AI.

For a decade, computer science was the golden major. It promised six-figure starting salaries, endless job openings, and a clear path from lecture hall to FAANG offer. That narrative is cracking. The same students who would have lined up for CS 101 are now enrolling in AI and machine learning tracks instead, drawn by a job market that rewards those skills at a steep premium. The shift is real, and it is reshaping higher education faster than universities can keep up.

Where are the students actually going?

The exception in the UC data is telling: UC San Diego's new AI major is booming. Nationally, MIT's "AI and decision-making" track, launched in 2022, is now its second-largest major, and dozens of US universities launched dedicated AI programs in the same window (TechCrunch, Feb 2026; NYT, Dec 2025). The shift is not subtle.

• UC San Diego AI major: booming while CS declines
• MIT "AI and decision-making": now the #2 major
• Dozens of new US AI programs launched in 2 years
• Overall college enrollment: up 2% — CS is the outlier

Students are reading the job market directly. Entry-level software roles tightened, internship pipelines narrowed, and the "AI is coming for junior devs" discourse did the rest. The degree that promised a golden ticket no longer looks like one — so students are chasing the flavor of compute that does.

What is China doing differently?

The China comparison is the part most US coverage misses. Roughly 60% of Chinese university students and faculty now use AI daily, Zhejiang has made AI coursework mandatory across its schools, and Tsinghua runs interdisciplinary AI colleges designed for exactly the students US universities are losing from CS (TechCrunch, Feb 2026).

The structural difference matters. China treats AI literacy as a national curriculum goal, not a departmental elective. US universities are still debating whether AI belongs in the major list at all. That gap compounds yearly — each cohort of Chinese students graduates with AI fluency baked into their thinking, while American counterparts have to patch it on later.

China's approach is not just about adding classes. Tsinghua's interdisciplinary AI colleges combine computer science, mathematics, and domain expertise — training students who can apply AI to healthcare, manufacturing, and climate science, not just build chatbots. The US produces AI specialists. China is producing AI-literate generalists across every field.

The US is debating whether to add AI courses. China made AI literacy a national curriculum goal. That gap compounds.

— Theo Okafor

Is the CS degree really dying?

No — the plain CS major is shrinking, not the field. The data shows a rotation: the same students who would have enrolled in CS are choosing AI-flavored degrees, data science tracks, or adjacent majors with a clearer AI hook. The total computing talent pool is holding steady; its label is changing (TechCrunch, Feb 2026).

That rotation has a cost. Traditional CS curricula teach systems, algorithms and fundamentals that AI degrees often skim. If the industry needs both depth and AI fluency, students who chase only the shiny new major could find themselves one tool-thin.

The CRA's latest survey confirms the trend across North America: 62% of computing departments reported CS enrollment declines, while AI and cybersecurity programs grew at 56-58% of institutions (CRA, 2026). The talent is not vanishing. It is moving to where the job postings point.

What does this mean for the AI job market?

For hiring, the shift cuts both ways. A wave of AI-major graduates will flood the market within three years, but employers already report that "AI skills" in a posting is not the same as shipped experience — senior roles want AI 3x more often than entry-level ones, and the gap is widening (Job Lobster, 2026).

The data tells a clear story: employers want AI skills, but they want them on top of solid engineering fundamentals, not instead of them. A graduate who can fine-tune a model but cannot debug a race condition is only half-hireable. The market rewards depth and breadth — not the shortcut.

The real risk is a flooded entry-level market in 2028-2029. Thousands of AI-major graduates will compete for the same junior roles, and employers will start filtering on shipped open-source projects and portfolio depth, not degree titles. The students who build while they learn will win. The ones who coast through an AI major without shipping anything will face the same bottleneck CS grads face today.

Can you get an AI job without a CS degree?

Yes — and the path is widening. AI-native startups care more about shipped projects and GitHub contributions than degree titles. Bootcamps, self-taught portfolios, and open-source contributions are landing roles at companies like Hugging Face, Mistral, and dozens of YC-backed firms. The key differentiator is not the diploma. It is demonstrable skill with LLMs, fine-tuning, and deployment.

That said, a CS degree still opens doors at larger enterprises and top research labs. Google, Meta, and OpenAI all require degrees for most research positions. The degree is not dead — it is being unbundled. Some students take the AI major and supplement with open-source work. Others skip college entirely and build in public. Both paths are viable; neither is guaranteed.

Which universities have the best AI bachelor's programs in 2026?

• MIT — "AI and Decision Making" (now #2 major on campus)
• UC San Diego — dedicated AI major (only UC with one)
• University of South Florida — AI & Cybersecurity college (3,000+ students)
• University at Buffalo — "AI and Society" department (7 new degrees)
• Carnegie Mellon — first AI bachelor's (2018), still the benchmark
• Northwestern, Columbia, USC — launching AI programs fall 2026

What is the bottom line?

CS enrollment fell 6% in a single year while AI majors boomed — and China's 60% daily AI usage makes the US rotation look slow. Students are not abandoning technology; they are betting on the part of it that seems to pay the most. Whether the plain CS degree or the AI major wins, the lesson is identical: follow the jobs, not the curriculum.

The students making this choice right now will define the AI workforce of 2030. The smart ones are not choosing between CS and AI — they are choosing both paths at once. A CS foundation with an AI concentration, open-source projects on the side, and shipped work beat either degree alone. The exodus is real. The destination is still being built. And the students who move first will have the advantage.

Sources and further reading

• TechCrunch: "The great computer science exodus" (Connie Loizos)
• NYT coverage of AI majors
• CRA: Generation CS enrollment data
• Which AI Skills Actually Get You Hired in 2026?
• AI Agents Hit the Payroll
• Laid Off and Rehired at $155k With AI
• The Best AI Models of 2026, Ranked by Real Users