US computer-science enrolment falls by double digits in preliminary 2026 data

Computer and Information Sciences enrolment has fallen by double digits in the National Student Clearinghouse's preliminary fall 2026 data: 12.4% at four-year undergraduate institutions, 12.3% at two-year undergraduate institutions and 10.4% at graduate institutions. Overall undergraduate enrolment increased 0.6%, while graduate enrolment declined 1.6%. The contrast is sharp enough to establish a field-specific shift. It is not enough to establish that artificial intelligence caused it.[1]

This is not a one-term anomaly. Final fall 2025 data showed declines across every award level and institution type, ranging up to 14% at graduate level. Spring 2026 then recorded falls of 8.4% at four-year institutions and 11.2% at two-year colleges. The preliminary fall 2026 release points in the same direction and at a greater rate than the wider undergraduate market.[1][2][3]

A descriptive dataset, not an explanation

The Clearinghouse data tracks enrolment categories. It does not ask students why they selected or rejected a major, measure their expectations about AI, or connect individual decisions to job advertisements, wages, visa rules, tuition, programme capacity or institutional reclassification. Any claim that students are abandoning computer science because coding assistants can generate software would therefore require different evidence.[1]

Several explanations would need to be tested separately, including labour-market expectations, changes in international graduate enrolment and movement between programme classifications. The report also records health-field growth of 2.0% to 6.2% across the reported levels. That contrast describes where enrolment changed; it does not establish why students made their choices.[1][3]

Computer and Information Sciences is a field classification, not a count of every student learning computational skills. Students moving into separately classified programmes can change its totals without leaving technical education. Establishing how much such movement explains would require programme-level transitions rather than the aggregate figures in this release.[2][3]

What the preliminary status changes

The fall release uses 7.92 million enrolments reported by 46% of the institutions that submit data, as of 24 September. The final report, scheduled for January 2027, will use more comprehensive coverage and may revise the estimates. The Clearinghouse says the preliminary series emphasises measures that have historically tracked final trends reliably, but a 12.4% early decline should still be reported as preliminary rather than converted into a final headcount.[1]

For universities, the useful follow-up is to examine applications, enrolment, retention and programme transfers separately. Those measures would distinguish fewer entrants from more students leaving or changing majors. The aggregate decline identifies a problem to investigate, but does not identify which response would address it.[1][2]

For employers and policymakers, the figures are a warning about the future pipeline, not a forecast of an immediate skills shortage. Enrolment changes take years to reach graduation, and labour supply also depends on completion, immigration, occupational switching and productivity. The sound conclusion is narrower and still consequential: traditional computer-science enrolment has now declined across multiple terms while overall undergraduate participation has been steadier. Explaining why is the next research task, not a result already contained in the chart.[1][2][3]

Report an error or suggest a correction