CS Enrollment Drops 8.1% as AI Sparks Third Major Bust

The Hottest College Major Hits a Wall

Computer science enrollment at U.S. four-year universities plunged 8.1% in fall 2025 — the largest single-year decline for any major since tracking began in 2020. The drop knocked CS from the 4th to the 6th most popular undergraduate major nationwide, signaling what venture capitalist Deedy Das and other industry observers are calling 'the beginning of the third great recession in computer science history.'

Das's post on X sparked widespread debate across the tech world. 'If you picked computer science during the last downturn, you're doing pretty well right now,' he wrote. 'I suspect history will repeat itself today.' The comment resonated deeply, racking up thousands of reposts and replies from developers, students, and hiring managers grappling with a rapidly shifting job market.

Key Takeaways

• CS enrollment at four-year U.S. universities fell 8.1% in fall 2025, the steepest single-year drop on record
• CS degrees grew roughly 5x between 2008 and 2024 — more than double the growth rate of any other major
• Nobel laureate Simon H. Johnson (MIT) says AI has 'largely eliminated' programming as a reliable career path
• The Atlantic has declared 'the computer science bubble is bursting'
• Industry veterans compare this to two previous CS busts: 1984–1994 and 2002–2007
• Despite the enrollment decline, demand for AI-specialized talent remains historically high

Five Times Growth, Then a Sudden Reversal

The numbers tell a dramatic story. According to a deep-dive report published by The Washington Post on April 13, U.S. four-year institutions awarded roughly five times as many computer science degrees in 2024 as they did in 2008. That growth rate was more than double the second-fastest-growing field, exercise science.

For over a decade, CS appeared to be an unstoppable juggernaut. Students flooded into the major chasing six-figure starting salaries at companies like Google, Meta, and Amazon. Universities scrambled to hire faculty, expand lab space, and launch new programs to meet seemingly insatiable demand.

Then the music stopped. The 8.1% enrollment decline represents not just a statistical blip but a psychological shift among incoming students and their families. The question that once had an obvious answer — 'What should I major in to guarantee a good career?' — suddenly has no clear response.

AI Is Rewriting the Rules of Software Employment

The proximate cause of this downturn is unmistakable: artificial intelligence. Unlike previous CS busts driven by dot-com crashes or hardware market corrections, this one stems from the very technology the field created.

Simon H. Johnson, the Nobel Prize-winning economist at MIT, recently made headlines by stating that AI has 'to a considerable extent eliminated' programming as a dependable career pathway. His remarks carry particular weight given MIT's status as one of the world's premier computer science institutions.

The logic is straightforward and unsettling for aspiring developers. Tools like GitHub Copilot, Claude, GPT-4, and Cursor can now generate, debug, and refactor code at speeds no human can match. Devin, the so-called 'AI software engineer' from Cognition Labs, promises to handle entire development workflows autonomously. Companies like Klarna have publicly boasted about cutting engineering headcount while maintaining output, thanks to AI tools.

For a prospective college freshman weighing a four-year commitment, these signals are hard to ignore. Why spend $100,000+ and four years learning to write code that a $20/month AI subscription can produce in seconds?

The Two Previous CS Busts Offer Crucial Context

History suggests the picture is more nuanced than the doom-and-gloom headlines imply. Computer science has weathered two major enrollment collapses before — and each time, the field emerged stronger.

The First Bust: 1984–1994

The late 1970s and early 1980s saw an explosion of interest in computing, driven by the personal computer revolution. Companies like Apple, IBM, and Commodore captured the public imagination, and students rushed into CS programs. But the market quickly became oversaturated. Hardware commoditization, the AI winter of the late 1980s, and a general economic slowdown combined to crater enrollment. CS degrees declined for nearly a decade.

Those who stuck with the field, however, were perfectly positioned when the World Wide Web emerged in the mid-1990s.

The Second Bust: 2002–2007

The dot-com crash of 2000–2001 devastated the tech industry and sent CS enrollment into freefall. Horror stories of mass layoffs at companies like Pets.com, Webvan, and countless other startups made headlines. Outsourcing fears — the idea that all programming jobs would move to India and China — added to the anxiety.

Students who enrolled in CS programs during this trough graduated into one of the greatest tech booms in history. The rise of smartphones, cloud computing, social media, and the mobile app economy created millions of high-paying jobs.

Why This Time Might Be Different — Or Not

Bears and bulls are locked in fierce debate over whether the current downturn follows the same playbook.

Arguments that this time is different:

• AI coding tools are improving at an exponential rate, unlike previous 'threats' to developer jobs
• Major tech companies are explicitly hiring fewer junior developers
• The technology displacing workers is endogenous — it comes from within CS itself
• AI capabilities are compounding, not plateauing, making future disruption likely to accelerate

Arguments that history will repeat:

• Every previous prediction of 'the end of programming' has proven wrong
• AI tools still require skilled humans to architect systems, evaluate outputs, and handle edge cases
• The demand for software is effectively infinite — cheaper production means more products get built
• Reduced enrollment today means reduced supply in 4 years, which typically drives salaries back up

Das himself hints at the optimistic interpretation. His observation that people who 'picked CS during the last downturn' are doing well today implies a contrarian bet: the best time to enter a field may be precisely when everyone else is leaving.

The Market Is Bifurcating, Not Collapsing

A closer look at hiring data reveals a more complex reality than 'CS is dead.' The job market for software engineers is bifurcating sharply.

At the top end, demand for AI/ML engineers, infrastructure specialists, and systems architects has never been higher. Companies like OpenAI, Anthropic, Google DeepMind, and xAI are engaged in a fierce talent war, offering compensation packages exceeding $500,000 for experienced researchers. NVIDIA, whose GPUs power the AI revolution, saw its market cap surpass $3 trillion partly on the strength of insatiable demand for AI infrastructure.

At the entry level, however, the picture is bleak. Traditional junior developer roles — the bread-and-butter first jobs for CS graduates — are disappearing or being dramatically restructured. Companies that once hired cohorts of 50 new-grad engineers are now hiring 10, armed with AI tools that multiply each developer's productivity.

This bifurcation creates a painful paradox for students: the field offers extraordinary opportunities at the top while simultaneously pulling the ladder away from the bottom.

What This Means for Students and the Industry

The implications of the CS enrollment decline ripple far beyond university campuses.

For current and prospective students:

• A CS degree alone is no longer a golden ticket — specialization in AI, security, or systems is increasingly essential
• Interdisciplinary skills (CS + biology, CS + finance, CS + law) may offer more durable career protection
• The contrarian play — entering CS when others are leaving — has historically paid off handsomely
• Understanding AI tools is becoming more valuable than raw coding ability

For the tech industry:

• A smaller pipeline of CS graduates in 4–5 years could create talent shortages, particularly if AI fails to fully replace human developers
• Companies relying on AI to replace junior engineers may face institutional knowledge gaps
• The 'missing generation' of developers could become a strategic liability

For universities:

• Programs must rapidly evolve curricula to emphasize AI literacy, systems thinking, and human-AI collaboration
• Schools that fail to adapt risk losing relevance as students seek alternative pathways like bootcamps and self-directed learning

Looking Ahead: A Field in Transformation

The current CS enrollment decline is real, significant, and driven by legitimate concerns about AI's impact on traditional programming careers. But declaring the death of computer science would be as premature as it was in 1990 or 2003.

The field is not dying — it is metamorphosing. The skills that defined a successful CS career in 2015 (writing clean code in Python or Java, building CRUD applications, maintaining legacy systems) are indeed being commoditized by AI. But the skills that will define success in 2030 — designing AI systems, ensuring algorithmic safety, building the infrastructure that powers intelligent applications — remain deeply technical and profoundly human.

The students who enroll in computer science programs today, when the crowds are thinning and the headlines are grim, may find themselves in the same enviable position as those who entered during the dot-com bust. They will graduate into a world that runs entirely on software and AI, with fewer competitors and more leverage.

As Deedy Das implied, the best time to plant a tree is when no one else is planting. The third great CS downturn may ultimately prove to be the third great CS buying opportunity — for those with the conviction to see it through.