For years, students interested in technology were given pretty straightforward advice: learn to code. Computer science degrees became incredibly attractive, software jobs paid well, and artificial intelligence only seemed likely to make technical skills even more valuable.
Now, something interesting may be happening. Students appear to be cooling on computer science at the same time AI is becoming increasingly capable of writing code itself.
New data from Nova Learning shows Computer Science and AI falling sharply as the primary academic interest among the middle and high school students it surveyed. The category accounted for 42.3 percent of responses in 2025, but only 27.9 percent in 2026, while Engineering nearly doubled from 11.9 percent to 23 percent.
The details are even more interesting than the headline numbers. Within Computer Science and AI, Software and Data Science suffered the steepest proportional decline, losing 44 percent of its 2025 share.
AI itself recorded the largest absolute decline of any individual subject in the survey, falling 6.2 percentage points. Cybersecurity and Human-Computer Interaction held up considerably better, suggesting students may be becoming more selective about which parts of computing still feel attractive.
In other words, students do not necessarily appear to be running away from technology. They may simply be becoming less interested in the traditional learn-to-code path.
Engineering tells a similar story from the opposite direction. Aerospace and Mechanical Engineering together accounted for more than 60 percent of students selecting Engineering in 2026 and were responsible for 81 percent of the category’s growth.
Electrical and Computer Engineering also gained ground. That makes this look less like a rejection of technical fields and more like a shift toward areas tied to hardware, machines, and physical systems.
Software development once looked like one of the safest technical careers a student could pursue. Today, AI tools can generate functions, debug code, build interfaces, explain unfamiliar codebases, and increasingly complete substantial chunks of software development with relatively little human input.
That does not mean programmers are disappearing, of course. It does mean the value proposition of simply knowing how to write code is changing quickly.
Students entering high school today are looking at a very different technology industry than someone choosing a major a decade ago. They see AI coding assistants getting better, technology companies announcing layoffs, and constant debate over how many entry-level programming jobs AI could eventually replace or reshape.
At the same time, aerospace, robotics, manufacturing, energy, and other engineering-heavy industries are getting enormous amounts of attention. Building aircraft, robots, chips, engines, medical devices, and physical infrastructure may simply feel more tangible than competing with an AI model that can produce software in seconds.
Nova’s numbers should still be treated cautiously. The organization surveyed 463 students in 2025 and 677 in 2026, primarily from the U.S. West Coast.
The sample was not designed to represent students nationally, and the two groups contained different students. Nova is measuring differences between two cohorts rather than watching the same students change their minds.
Still, Nova’s findings do not exist in isolation. The National Student Clearinghouse Research Center found that undergraduate enrollment in Computer and Information Science programs at four-year institutions fell 8.1 percent in fall 2025, dropping to roughly 606,000 students.
That broader decline makes Nova’s smaller survey harder to dismiss as an oddity. It also suggests the cooling interest in computer science may extend beyond one education company and one set of younger students.
None of this proves AI is causing students to abandon computer science. Nova’s own data cannot establish why students made their selections, and plenty of other economic and cultural factors could be involved.
The timing is still hard to ignore. For years, the message was that everyone should learn to code because software was eating the world, but now software can increasingly help write itself.
The next generation of technically minded students may already be adjusting accordingly. If that continues, computer science may no longer be the automatic default for students who simply know they want a future in technology.
Support independent tech journalism
NERDS.xyz is independently owned and operated. If you enjoy my coverage of Linux, AI, hardware, cybersecurity, and tech culture, consider supporting the site on Ko-fi.
Support NERDS.xyz
