The Pulse of Higher Ed

By Kathleen Ives

When the New York Times ran a piece in August pointing out that companies are throwing billions into AI with little to show for it, I had déjà vu. It took me right back to the late 1990s, when everyone thought the web would change everything overnight. It did change everything—but not before years of hype, broken promises, and a lot of failed businesses. The story feels familiar, but AI is moving at a very different speed.

The adoption curve for AI is off the charts. ChatGPT hit 100 million users in two months—faster than TikTok or Instagram ever did (Chow, 2023; Hu, 2023). Almost every company you talk to is running AI pilots. In customer service and coding, we’re already seeing measurable gains. It feels like the web boom, but on fast-forward.

But here’s the thing: adoption is not the same as payoff. According to Lohr (2025), about eight in ten companies are using generative AI. At the same time, just as many say they haven’t seen a real bottom-line impact. A separate report showed that 42% of companies abandoned most of their AI projects in 2024–25—up sharply from 17% the year before (Wilkinson, 2025). And a recent MIT report put the number even higher, finding that as many as 95% of corporate generative AI pilots are failing to deliver meaningful results (Estrada, 2025). The U.S. Bureau of Labor Statistics did record a 2.4% productivity jump in Q2 2025, but it’s hardly the kind of step-change that signals a true revolution (U.S. Bureau of Labor Statistics, 2025).

If you lived through the dot-com era, this pattern looks familiar. Companies rushed online, but most didn’t have the models, infrastructure, or customer base to make it work. A lot of them flamed out. But a handful of survivors—Amazon, Google, eBay—figured out how to translate clicks into profits once the infrastructure caught up. That’s where we are with AI: massive experimentation, high failure rates, and a lot of lessons still being learned.

Here’s how the parallels stack up:

Now, what does all this mean for higher education? This is where things get messy. Our sector is facing financial and political headwinds that make big bets on AI much harder. Moody’s downgraded the outlook for higher ed to negative earlier this year (Moody, 2025a). Tuition discounting keeps climbing, squeezing budgets even further (Moody, 2025a). And on top of all that, the Trump administration has gutted the Department of Education staff, creating instability in areas like financial aid and compliance (Katz, 2025). In this climate, large-scale AI adoption isn’t just unlikely—it’s nearly impossible for most institutions.

But that doesn’t mean AI has no role in higher ed—quite the opposite. The smart play right now is targeted, practical adoption—the kinds of tools that ease pressure on people and budgets rather than adding to them. Think about AI that can speed up advising, help draft grant proposals, or take some of the routine work off IT help desks. These aren’t the splashy, campus-wide transformations we read about in headlines. Still, they’re the projects with quick payback, minimal compliance risk, and the ability to free up time and resources for the work that really matters.

So where do we go from here? For colleges and universities, the takeaway is pretty simple: start small, be intentional, and keep the mission front and center. Pick one or two areas where AI can make an immediate difference—cutting wait times for students, streamlining paperwork, or reducing repetitive tasks. Then track the impact, share what works, and build from the wins. What higher ed can’t afford right now is chasing shiny tools just to claim we’re “doing AI.” The point isn’t to be first out of the gate—it’s to make smart, lasting choices that protect mission and margin in a tough political and financial climate.

Kathleen Ives is UPCEA’s Chief Business Development Officer & Senior Vice President of Member Engagement. She oversees membership, partnerships, research, and consulting, leveraging her cross-sector experience to drive engagement, build alliances, and advance UPCEA’s mission.

References

Chow, A. R. (2023, February 2). ChatGPT is the fastest-growing app of all time. Time. https://time.com/6253615/chatgpt-fastest-growing/

Estrada, S. (2025, August 18). MIT report: 95 percent of generative AI pilots at companies failing. Fortune. https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/

Hu, K. (2023, February 1). ChatGPT sets record for fastest-growing user base. Reuters. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/

Katz, E. (2025, July 14). Education Dept. can proceed with mass layoffs after Supreme Court ruling. GovExec. https://www.govexec.com/workforce/2025/07/education-dept-can-proceed-mass-layoffs-after-supreme-court-ruling/406714/

Lohr, S. (2025, August 13). Companies are pouring billions into A.I. It has yet to pay off. The New York Times. https://www.nytimes.com/2025/08/13/business/ai-business-payoff-lags.html

Moody, J. (2025a, March 19). Moody’s downgrades higher ed outlook to negative. Inside Higher Ed. https://www.insidehighered.com/news/business/financial-health/2025/03/19/moodys-downgrades-sector-outlook-negative

Moody, J. (2025b, June 24). Tuition discounting hits another high. Inside Higher Ed. https://www.insidehighered.com/news/business/revenue-strategies/2025/06/24/tuition-discounting-hits-another-high

U.S. Bureau of Labor Statistics. (2025, August 7). Productivity up 2.4 percent in second quarter of 2025. https://www.bls.gov/news.release/pdf/prod2.pdf

Wilkinson, L. (2025, July 17). 42% of companies abandoned most AI projects, up from 17% in 2024. CIO Dive. https://www.ciodive.com/news/AI-project-fail-data-SPGlobal/742590/