For a university, AI visibility is whether ChatGPT, Gemini or Copilot names your institution, describes it accurately, and includes it when a prospective student asks for a recommendation, not whether your admissions office uses AI internally. Nearly half of high school students researching colleges are now putting that question to an AI assistant before they ever fill out an inquiry form, according to EAB's own February 2026 research, and what the assistant says back is shaped by sources most university marketing teams have never audited. Here is what the data actually shows and what a communications team can do about it.
How many prospective students are actually asking AI about colleges?
EAB, the higher-ed research and consulting firm, surveyed more than 5,000 high school students nationwide in October and November 2025 and published the results on February 25, 2026. It found that 46% of students now use AI tools such as ChatGPT during their college search, up sharply from 26% in spring 2025. That is close to a doubling in under a year, inside a search process that, for most of the last two decades, ran almost entirely through a university's own website and a handful of ranking sites.
EAB's companion parent survey, released separately in 2026, found that 57% of parents have used AI tools such as ChatGPT to compare colleges on their child's behalf. A university that only thinks about the student as its AI-visibility audience is missing more than half of the other decision-maker in the room.
Source: EAB, surveyed October to November 2025, published February 25, 2026, linked in the article.
Does what the AI says actually change anyone's decision?
Yes, measurably. EAB's student survey found that 18% of students, nearly one in five, removed a college from consideration specifically because of information an AI assistant surfaced about it. That is not a hypothetical reputational risk. It is a one-in-five chance, for every prospective student who asks, that whatever ChatGPT says about your institution costs you an applicant before your admissions team ever gets a name and an email address.
A separate 2026 survey from OHO, a higher-ed marketing firm, fielded in March 2026 among 530 respondents including 327 prospective and current undergraduates aged 17 to 22, found a similar split on the upside: 74% said AI helped them discover a school they hadn't previously considered, and 50% said AI helped them remove a school from consideration. The same survey found 36% of that age group now name AI as their first source in a college search, ahead of the 32% who start with Google, and 65% rate AI-generated information about colleges as accurate, up from 51% two years earlier. Trust in the answer is rising faster than most universities' effort to shape it.
Where is the AI actually getting its information about your university from?
The same places it gets information about any brand: your own site if it is crawlable and current, independent coverage in news and higher-ed trade press, forums like Reddit and College Confidential where applicants and current students compare notes, third-party ranking sites, and review platforms. An AI assistant's answer is a synthesis of whatever it can find across that set, weighted toward sources it judges credible and recent, not a lookup against an official record a university controls.
That is why a university's own admissions pages, financial aid pages, and program pages need the same basic hygiene any other brand's site needs for AI visibility: clear, current, uncomplicated prose stating tuition, acceptance rates, program details and outcomes in plain sentences a model can quote directly, not buried in PDFs, interactive calculators or JavaScript-rendered widgets a crawler cannot read. A program page that only states its admission rate through a dynamic chart has effectively told every AI crawler that it has no admission rate.
What role does earned media and third-party coverage actually play here?
A larger one than most university marketing plans currently budget for. Independent coverage, a well-sourced story about your research output, a faculty expert quoted on a relevant news story, a program covered in a higher-ed trade outlet, carries more weight with an AI system than your own admissions copy, for the same reason it carries more weight with a human reader: it was not written by the subject about itself. A university with faculty regularly quoted in national or trade press, and with research findings that get independent coverage rather than only a university press release, is feeding the exact kind of third-party signal an AI system treats as more credible.
This is also where student and alumni voice matters in a way a glossy brochure cannot replace. OHO's finding that AI-generated information is now trusted by 65% of prospective students, up from 51%, sits alongside forums and review sites where actual student experience gets discussed candidly. The same dynamic shows up with customer reviews for any other brand category: an AI system treats unprompted third-party commentary as a stronger signal than a marketing page, and a university's reputation online is increasingly built from the sum of what students and faculty say about it elsewhere, not from what the admissions office publishes.
Does this replace a university's existing SEO and enrollment marketing?
No, it sits alongside it and increasingly determines whether that other work gets seen at all. Search engine optimization still governs whether your pages rank in a traditional results page, but a rising share of a prospective student's research now happens inside an AI assistant's answer rather than on a results page with ten blue links to click through. A university that keeps its enrollment marketing budget pointed entirely at paid search and SEO, with nothing allocated to checking what ChatGPT, Gemini and Copilot currently say about it, is optimizing for a smaller and smaller share of how prospective students actually arrive at a decision.
What should a university's communications team actually do first?
Start by asking the four major assistants, ChatGPT, Gemini, Copilot and Perplexity, what they say about your university's admission rate, tuition, flagship programs and outcomes, and write down exactly where each answer is wrong, outdated or missing. Then work backward from each gap to its likely source: a stale tuition page, a program with no recent independent coverage, a department with no faculty voice in the press at all. A company newsroom page built for both reporters and crawlers does the same job for a university that it does for any other organization: it is the plain-text, dated, crawlable record an AI system can actually quote, and most university sites bury the equivalent of it inside a PDF-heavy press room nobody has updated since the last academic year.
The direct answer, for anyone who read this far
Nearly half of prospective students now ask an AI assistant about your university before they ever talk to your admissions team, and close to one in five have already dropped a school from consideration based on what that assistant told them. The fix runs through the same two levers that drive AI visibility for any brand: a crawlable site with plain, current facts, and genuine third-party coverage an AI system can treat as credible. If you want help auditing what AI assistants currently say about your institution, get in touch with our team.
