The University of North Carolina System made news last month when its Board of Governors approved a pilot program to offer accelerated three-year bachelor’s degrees across its 16 public universities. The headline framing was familiar: college costs too much, takes too long, and leaves students saddled with debt before they’ve earned their first paycheck. Compress the timeline, reduce the credits from 120 to 90, get students into the workforce faster. Problem solved.
Except the story underneath is considerably more complicated — and considerably more important.
The Ground Is Shifting Under the Degree Itself
While universities debate how many years a degree should take, AI is quietly redrawing the map of what entry-level work looks like. Stanford research confirms a 13% decline in entry-level hiring for AI-exposed roles since the rise of generative AI. The biggest tech firms cut early-career hiring by 25% from 2023 to 2024 alone. This is not a blip.
The mechanism matters. The tasks that once defined entry-level jobs — summarizing meetings, drafting memos, cleaning data, doing basic research — are precisely the tasks AI now handles most efficiently. Which means the learning curve that used to be built into those jobs has been automated away. New graduates are entering a labor market where the traditional on-ramp no longer exists, sandwiched between AI agents doing the routine work and senior employees who expect mid-level output from day one.
At the same time, the jobs that remain are demanding more. According to a Handshake report from April 2026, entry-level postings calling for AI skills nearly doubled year-over-year. Skills for AI-exposed roles are evolving 66% faster than other occupations. Employers are, in effect, asking for the judgment of experienced workers at entry-level salaries, with the expectation that AI will close the gap. Whether that expectation is realistic is another question.
Two Defensible Positions
Here is where it gets genuinely interesting — because there are two coherent ways to read this situation, and they point in opposite directions.
The case for accelerated degrees: If AI is handling more of the baseline cognitive work, then perhaps the knowledge threshold required to enter the workforce is functionally lower. The argument goes that much of the traditional four-year degree was always credential theater — general education requirements that served institutional inertia as much as student development. If students can reach workforce readiness in three years, and emerge with less debt, the compressed model is a rational adaptation to a changed environment.
The case against: In a world where AI performs routine and procedural tasks with increasing competence, the premium shifts decisively to what AI cannot do — ethical reasoning, creative synthesis, cross-disciplinary thinking, adaptability under ambiguity. These capacities are not built in major-specific coursework. They are built in the electives, the seminars, the humanities requirements, the unstructured intellectual exploration that tends to be the first casualty of curriculum compression. On this reading, cutting broad education is exactly the wrong response to an AI-saturated world. It optimizes graduates for the jobs AI is already eliminating.
The critics at Inside Higher Ed made this point plainly: a 90-credit degree redesigns education to produce humans who do surface-level work in fields that are demonstrably contracting. That is not workforce readiness. That is a race to the bottom.
The Policy Layer
Federal incentives are complicating this further. Beginning July 2026, expanded Workforce Pell grants will fund short-term credential programs — creating direct financial pressure on universities to demonstrate measurable job alignment in their curricula. The signal from Washington is clear: prove your education produces employment outcomes, or lose funding. Combine that with tuition-pressured families demanding faster degrees, and institutional administrators face a powerful convergence of incentives to compress and vocationalize.
None of those incentives are aligned with producing graduates who can think across disciplines in an economy being restructured by AI. The market signal and the actual workforce need may be pointing in opposite directions — and universities are being rewarded for following the market signal.
Meanwhile, bachelor’s programs explicitly in AI grew 114% from 2024 to 2025, and many institutions are working to embed AI literacy across all majors rather than treat it as a standalone subject. That is a more coherent response to the moment. But it is also more expensive and harder to compress into three years.
The Open Question
UNC System President Peter Hans has framed the three-year degree as an access initiative — a way to bring higher education within reach of adult learners, working students, and those who cannot afford a fourth year. That is a legitimate and important goal. Affordability and access in higher education are real crises, not manufactured ones.
But the mechanism matters enormously. Reducing cost by compressing time is not the same as reducing cost by rethinking funding structures, pricing models, or institutional overhead. One changes what education produces. The other changes how it is financed.
The deeper question the UNC pilot forces into view is one that higher education has not yet answered honestly: What is a college degree actually for in 2026? If it is a credential that signals baseline competence for an employer, then three years may be sufficient, and AI has simply lowered the bar. If it is the primary social institution for developing the kind of judgment and adaptability that an AI-saturated economy will increasingly demand — and reward — then we may be about to make a historic mistake in the name of affordability.
Are universities adapting to AI? Or are they adapting themselves out of their own purpose?
That is the question worth asking. The answer will shape the workforce — and the graduates — of the next decade.