Artificial intelligence isn’t dismantling higher education. It’s exposing what students have been showing up for all along.

by Tim Leogrande, BSIT, MSCP, Ed.S.

🗓  MAY 28 2026 • 8 MIN READ

To many college students, the value of a four-year degree isn’t demonstrated by the curriculum but by what it signals to hiring managers, colleagues, loan officers, and potential business or romantic partners. Baby Boomers built their entire lives around this concept; Gen X borrowed heavily against it; and Millennials — often grudgingly — treated it as the price of admission to the middle class.

I began thinking about this signaling after a conversation at a Memorial Day cookout with Haley, a sophomore at a $90,000-per-year, highly selective private liberal arts college. Between bites of a Coney Island hot dog and sips of Yuengling, Haley casually shared that she had the Claude chatbot “write basically every assignment so far” during her college career. She said she still went to class, she still liked her professors and classmates, and she still planned to graduate. When I asked Haley why she was outsourcing an education valued at $360,000 to a chatbot, she suddenly stopped chewing and looked at me as though the question itself were outdated. “For the degree and connections,” she deadpanned. “Obviously.”

At a time when more than 40% of recent graduates are working in jobs that don’t typically require a college degree, a trend that has held steady for about a decade, surveys reveal intense public skepticism about the value of higher education, even as most current students still say their own degree is worth it. It’s becoming increasingly apparent that college’s value proposition now has less to do with the transmission of knowledge than many admissions officers are willing to admit.

<aside> 💡

Most students pursue a degree to credibly signal their intelligence, work ethic, leadership potential, and reliability. Even as AI reshapes how we teach, the American higher education system will endure for three reasons. First, performing well in school is a way for ambitious students to distinguish themselves. Second, college provides an opportunity to establish valuable personal and professional connections. Third, many employers continue to view a diploma as evidence that a candidate is more capable and worth hiring.

</aside>

A student at a state university who pays $12,000 per year in tuition and a Yale student who pays $68,000 are increasingly receiving access to the same AI tutoring assistance when they sit down to write a paper. But the Yale student is paying for something else. In truth, she always has been. Now she’s just more aware of it.

When we spoke, Haley had already come to terms with this reality. What her tuition is purchasing is not simply instruction but the legitimacy conferred by the diploma itself, as well as the valuable networking opportunities that come with it. That includes peers and study partners, faculty connections, alumni networks, and dinner-party introductions ten years after graduation. AI can’t duplicate any of these things, and it’s not trying to.

Admittedly, this signaling concept has always had a problem. If a diploma is purely a credential, then why do employers continue to prefer graduates from schools with more rigorous curricula? Clearly, something is being gained. Part of it, perhaps, is the ability to pay attention, a tolerance for tedium, and the discipline to sit in a room and finish something. Whether AI erodes these habits or merely exposes that universities were never the only place to acquire them is the question administrators should be losing sleep over.

<aside> 💡

The real disruption within higher education is not that students can cheat using AI. Students have always cheated. It’s that a significant part of what students pay tuition for is now available instantly and at almost no cost from a chatbot that, on certain tasks, can rival the instruction provided by an adjunct.

</aside>

This argument is strongest within disciplines where the deliverable is a written artifact and the knowledge being tested is interpretive: literature, history, the social sciences, and the humanities. The case is much weaker in nursing, engineering, the lab sciences, and the trades, where a credential often represents not just classroom instruction but hours of supervised, in-person practicums that no chatbot can perform on a student's behalf.

A winner-take-all dynamic is already taking shape, though AI is only one factor. The demographic and enrollment cliffs that higher education has been bracing for since the late 2010s would have closed hundreds of small schools regardless of what products OpenAI or Anthropic released. Artificial intelligence increases that pressure, but it doesn’t create it.

Flagship state schools, respected national universities, and elite private colleges will continue to thrive thanks to their financial, cultural, and reputational advantages, while other less well-known institutions will be forced to close their doors. Haley’s college will survive this winnowing, but many won't. The schools that do close will be the ones that were already struggling to articulate why a student should choose them over a credential earned faster, cheaper, or from a chatbot.

Either outcome could end up being a net positive, but the assumption that many readers of this blog were raised with — that everyone should attend college to pursue knowledge, improve their research and writing abilities, and sharpen critical thinking skills for their own sake, rather than for upward mobility — may no longer be relevant.

Whether four-year schools can continue to persuade potential students that there is value in classroom instruction, completing assignments without AI assistance, and debating a flawed thesis statement with human faculty will be crucial to the future of higher education. But they will be pitching these ideas to a generation of people who may ask what the Krebs cycle has to do with landing a good job, and who are already very comfortable with using an LLM to assist with their schoolwork.

The individuals who are most likely to be receptive to this pitch come from affluent, highly educated families. These students may still place deep trust in institutional reputation and be willing to sacrifice some short-term efficiency for the reassurance of a formal credential they will carry for life. Yet even within this demographic, that premise may already be weakening.

A 2024 Pew survey found that 26% of U.S. teens had used ChatGPT for schoolwork, double the share from the previous year. By early 2026, Pew put that figure at 54% across all AI chatbots, with 10% of teens admitting they let artificial intelligence do “all or most” of their schoolwork. In other words, the cultural resistance many schools hope to preserve is already eroding.