After nearly two decades of rapid growth, computer science enrollment is declining across U.S. colleges just as AI systems become increasingly capable of writing software — raising an important question about whether students are beginning to rethink the traditional path into technology careers.
WHAT’S HAPPENING
Computer science has been one of higher education’s biggest growth stories for nearly two decades.
That trend has now reversed.
New data from the National Student Clearinghouse Research Center shows undergraduate enrollment in Computer and Information Sciences fell across every major institutional sector in spring 2026.
Enrollment declined:
8.4% at four-year institutions
9.3% at primarily associate-degree granting baccalaureate institutions
11.2% at two-year colleges
The decline stands out because overall undergraduate enrollment increased 1.3% during the same period. (National Student Clearinghouse)
Separate research from Stanford economist Jacob Light found computer science enrollment declined during the 2025–26 academic year for the first time in roughly two decades.
Light analyzed more than 50 million course sections across 1,019 U.S. colleges and universities and found computer science course enrollment at the average institution in his sample fell about 4.6% year over year. (Yahoo Finance)
The timing overlaps with the rapid rise of generative AI and increasingly capable coding assistants.
But there is an important distinction:
The evidence does not prove AI caused the decline.
WHY IT MATTERS
For years, computer science offered students a relatively clear proposition:
Learn to code.
Earn a computer science degree.
Enter a rapidly growing and often highly paid technology industry.
AI is complicating that equation.
Coding assistants can now generate software, debug programs, explain unfamiliar code and perform tasks that previously required considerably more manual programming.
At the same time, the entry-level technology job market has become more difficult.
Students deciding what to study are therefore looking at a different future than students entering college five or ten years ago.
That does not mean computer science is becoming obsolete.
But it may mean the value proposition is changing.
The future engineer may need less emphasis on manually producing every line of code and greater skill in areas such as system design, AI orchestration, cybersecurity, architecture, data, problem definition and understanding how software fits into a larger organization.
WHO BENEFITS
Students who combine computer science with AI skills could be well positioned if software development shifts toward managing and directing increasingly capable AI systems.
Universities that modernize computer science programs may benefit by teaching students how software development is actually evolving rather than preparing them exclusively for traditional programming workflows.
Related disciplines such as data science, cybersecurity and AI engineering could gain students who still want technology careers but are reconsidering conventional computer science programs.
Employers could eventually gain graduates trained for a more AI-native technology environment.
WHO LOSES
Computer science programs that fail to adapt could face additional enrollment pressure if students begin questioning whether traditional curricula match the emerging job market.
Entry-level developers focused primarily on routine coding tasks may face greater competition as AI handles more implementation work.
Technology companies could also face a longer-term problem if declining enrollment eventually reduces the pipeline of people with deep computer science foundations.
And students could lose if they interpret the trend too aggressively.
AI may change software engineering significantly without eliminating the need for people who understand algorithms, systems, computing infrastructure and software architecture.
WHAT HAPPENS NEXT
The next several college cohorts will tell us whether this is a temporary correction after years of extraordinary growth or the beginning of a more significant change in how students view computer science.
There are already reasons to watch closely.
National Student Clearinghouse data also shows computer science students had the lowest second-year persistence rate among the ten largest bachelor’s-degree fields, at 85%, compared with 93.1% for engineering. (National Student Clearinghouse)
But AI is only one possible explanation.
Light identified several potential factors, including a weaker entry-level software job market, uncertainty about the future value of computer science degrees, changes in student preparation and demographics, and competition from growing fields such as data science. His research cannot determine which factor is responsible. (Yahoo Finance)
That caution matters.
The signal is not:
AI killed the computer science degree.
It hasn’t.
The signal is that one of higher education’s most powerful growth trends has suddenly changed direction at exactly the moment when AI is changing what it means to write software.
Students make education decisions based partly on what they believe the future will look like.
And they may not have to wait for AI to completely transform the technology job market before they begin changing what they choose to study.
