Computer science enrollment is plunging, but AI is taking over the rest of the college campus as it reshapes how students learn and work

Undergraduate enrollment in computer and information sciences at four-year colleges fell 8.4% in spring 2026 vs. a year earlier, after a 3.6% fall in fall 2025, with graduate enrollment down 14% over the same period. Universities including Colby, the University of Florida, and Ohio State are responding by expanding cross-disciplinary AI education—e.g., Colby’s Davis Institute ($30M gift in 2021) runs an eight-week “Interdisciplinary Summer Bridges” program with 30 students across 23 majors. Overall, the article suggests demand is shifting from single-discipline computer science into broader AI fluency rather than disappearing, with limited immediate market impact.

Analysis

The real market mechanism here is not a collapse in tech demand; it is a re-mix of the labor stack. If AI literacy becomes a baseline across majors, the marginal buyer shifts from “hire more coders” to “buy software that lets non-coders automate workflows,” which is mildly supportive for enterprise software and services vendors with governance/integration capability. IBM is a modest beneficiary because its monetization path is implementation, training, and control layers rather than frontier-model hype; that makes the revenue path slower but stickier than pure AI infrastructure bets.

The second-order losers are the labor-arbitrage model and any education/training business exposed to a narrow coding narrative. As the labor market rewards domain-specific AI operators over generic CS degrees, offshore dev shops, legacy staffing, and low-end IT services should face margin pressure over the next 6-18 months as pricing power shifts toward productivity tools. The immediate data point is weak, but the larger risk is a 2027 hiring cycle where entry-level software demand stays soft even if overall AI adoption rises.

Contrarian view: the consensus may be too bearish on tech talent and too bullish on “AI replacing coders.” A broader AI curriculum widens the addressable market for enterprise software and professional services, while reducing dependency on scarce specialist hires. What would falsify that thesis is re-acceleration in entry-level CS hiring or evidence that universities are adding AI classes without any incremental enterprise spending on training, consulting, or compliance.

Key Decisions for Investors

• Small tactical long IBM on pullbacks over the next 1-3 months; thesis is incremental demand for AI implementation, training, and governance, with low direct revenue sensitivity but better durability than pure AI hype names. Falsify if services/consulting bookings or margins fail to inflect over the next two quarters.
• Do not short education/edtech on this print alone; the signal is more about curriculum substitution than demand destruction. Keep a watchlist on operators with narrow coding-bootcamp exposure, but wait for evidence of weaker enrollment or pricing before acting.
• Relative-value long IBM vs. a broad software beta basket over 6-12 months if you want AI monetization with less multiple risk; the setup favors firms that sell into enterprise workflow adoption, not just model enthusiasm. Trim if IBM starts underperforming on bookings or if AI-related services cease to grow faster than core revenue.
• Set a catalyst alert for university-to-employer hiring data over the next 2 hiring cycles: if internship and entry-level software requisitions remain weak, the labor-supply thesis becomes a 2027 margin headwind for IT services and staffing.