The Pulse of Higher Ed

By Bruce Etter

During arms-control talks in the 1980s Ronald Reagan made “trust, but verify” one of the most quoted phrases of the era. Four decades later, prospective online and continuing education students appear to have quietly borrowed it back and pointed it directly at their search bar. Trust, but click.

That is the story of AI Search in Higher Education: How Prospects Search in 2026, conducted in partnership between UPCEA and Search Influence. The headlines sound contradictory: prospects are using AI search more than ever, trusting it more than ever, and questioning it more than ever, all at the same time. Look closer and the contradiction disappears. A more sophisticated AI consumer is emerging in the online and professional continuing education market. Prospects are happy to accept AI’s opening argument. They just insist on seeing the evidence before they render a verdict.

Start with the use side of the ledger. Daily use of AI-powered search tools jumped 15 percentage points in a single year, from 24% in 2025 to 39% in 2026. The share of prospects who never use AI search fell from 17% to 12%. Fifty-eight percent now treat AI tools the way they would treat a traditional search engine, up from 50% a year ago. Adoption is no longer the question. The tools have moved into the routine.

Trust is the more interesting story, and it is moving in the same direction. Eighty-three percent of prospects read AI-generated overviews when they appear inside a Google or Bing result, up slightly from 79%. Among those readers, 64% say that being cited inside an overview makes them trust the website more, up from 56%. Fifty-seven percent of prospects report that their trust in AI Overviews has increased in the past year; only 12% say it has declined. Use and trust are not the same thing, but in 2026 they are pulling in the same direction.

That trust curve should get the attention of any institution trying to reach the modern learner. AI did not just move into the search stack. It moved in, took off its coat, and got comfortable. Higher education understands the language of citation better than most industries, and this is a version of it: being named as a source inside an AI Overview functions much the way a peer citation functions in a piece of scholarship. It confers perceived standing. And a growing majority of prospective students respond to it accordingly.

Even as trust has risen, so has scrutiny, though not quite in the way the headline number suggests. In 2025, roughly a third of prospects (32%) reported no concerns whatsoever about AI search tools. In 2026, that share collapsed to 14%. The rise in “at least one concern” from 68% to 86% is not because any single fear went vertical; the top-cited concerns, validity and accuracy, actually ticked down slightly year over year. What changed is that fewer prospects are willing to say the technology raises no concerns at all. In parallel, verification became near-universal: 59% of overview readers click through to cited sources, and only 1% say they never do.

Read those numbers together and a very clear behavior emerges. Prospects are embracing AI and wisely not blindly accepting it as gospel either. They are doing what smart researchers do, just at a different scale: read the summary, then check the sources. AI provides the abstract. The linked website provides the evidence. The trust and the skepticism are not in conflict. They are two halves of the same verification loop.

The consequence for PCE marketing is that your program page now serves two visitors at nearly the same moment: an AI system deciding whether the page is authoritative enough to cite, and a prospective student who read the overview, saw your name inside it, and clicked through to see whether the AI got it right. Both want substance: concrete, verifiable detail on what the program is, who it is for, how long it takes, how much it costs, and what graduates go on to do. And it is worth remembering that your program page is not the only surface AI reads on your behalf. Third-party profiles, alumni listings, review sites, and social platforms all feed the answer AI eventually assembles, and inconsistency across those surfaces can keep otherwise-qualified programs out of the citation.

That consideration set has never been more precious. Fifty-six percent of prospects seriously consider three or fewer PCE programs before deciding where to apply, and 7% consider just one. Among prospects at least somewhat likely to use AI in program research, half (50%) would use it specifically to narrow their consideration set and 35% would let AI pick the best program for them outright. Another 56% of that same AI-open group would use AI to do additional research on programs they are already interested in, meaning AI is joining the deep dive, not replacing it, and 72% are more likely to consider a program that AI features prominently in its results. AI is being invited into the earliest, most consequential moment of the funnel, the moment a prospect goes from “everyone” to “these three.” Miss on either side of the citation, and the door quietly closes before an inquiry is ever made.

The 2026 findings point to a fairly disciplined to-do list. Treat inclusion in AI Overviews the way you already treat organic ranking: earned, engineerable, and now a direct trust and traffic driver. Write program pages worth citing, in plain English, structured for machines, and backed by specific, checkable details. Audit every surface AI reads on your behalf so the story stays consistent across them. Remember that traditional search still starts the journey for 80% of prospects and university websites remain the most trusted source in the mix (74%); AI supplements those behaviors, it does not replace them. And keep an eye on video: 67% of prospects search YouTube the way they would search Google.

Above all, resist the temptation to read “86% have concerns about AI” as a warning that adoption will slow down. It will not. What is happening is that a discerning user base is emerging, one that will keep asking AI to do more of the heavy lifting while also keeping receipts. Institutions that give both the AI system and the prospect something worth trusting will move into that shrinking consideration set. Institutions that do not will find themselves outside a door that was closed by an AI Overview they were never in.

Trust, but click. Prospects have already learned the discipline. Your program pages, and every other surface AI reads on your behalf, need to be worth both halves.

Bruce Etter serves as the Senior Director of Research and Consulting at UPCEA, where he leads the development and management of research initiatives for UPCEA’s Research and Consulting division and its clients.