Ask your communications team about your institution’s reputation and they can probably give you a decent answer. They know the brand research, the media coverage, the enrollment narrative, and what shows up on Google.
Now ask a different question: What does ChatGPT think of us?
Ask what your institution is known for, whether it’s a good value, whether graduates get good jobs, why enrollment is declining, or how it compares with the college down the road. Ask Perplexity and Gemini, too.
You may not recognize the institution they describe.
That’s because AI doesn’t care about your brand platform, your org chart, or the messaging framework you spent nine months developing. It sees the ecosystem around your institution: your website, earned media, government data, rankings, Reddit conversations, accreditation reports, old news stories, faculty expertise, student complaints, employment outcomes, and a whole lot more.
Then it connects the dots. Sometimes it connects dots you’d rather it didn’t (have you noticed your school’s “AI Overview” in a Google search?).
You can’t prompt-engineer your way out of a reputation problem
There is plenty of good work happening around making institutional content more visible to generative AI. Colleges should absolutely understand how AI platforms find, interpret, and cite information.
But treating this as the next iteration of SEO misses the bigger issue.
Suppose your brand promises career readiness, but employment outcomes are nearly impossible to find. You describe the institution as innovative, but your academic program pages haven’t changed meaningfully in six years. Leadership talks about transformation, while years of news coverage document enrollment declines, budget cuts, and turnover.
You might have an AI visibility problem. You might also have a credibility problem.
The interesting question isn’t, “How do we make ChatGPT say better things about us?” It’s “Why is ChatGPT saying this about us in the first place?”
That is a much more useful conversation.
And it changes how we should think about the content institutions produce. Prospective students increasingly don’t need to navigate your website, download your materials, or consume your carefully constructed recruitment journey to learn about you. They can simply ask.
Your next viewbook isn’t for students. It’s for AI.
Okay, it’s still for students. But the larger point stands. Institutional content now has another job: it has to contribute clear, credible, current evidence to an information ecosystem increasingly interpreted by machines.
TIP: A helpful how-to guide, of sorts.
That doesn’t mean stuffing websites with language designed for AI. It means making sure the evidence supporting your institutional story actually exists and can be found. The outcomes buried on page 37 of a PDF, the compelling faculty story that never escaped the campus newsroom, and the remarkable employer partnership everyone on campus knows about but nobody outside it does aren’t just missed marketing opportunities anymore. They’re missing reputation signals.
SPOILER ALERT: This overlaps in many ways with your ACA Title II work (already?) underway.
Run an AI Reputation Audit
Forget the technical prompts for a moment and ask the questions prospective students, parents, employees, donors, journalists, and community members might actually ask.
Is this a good college? What is it known for? Is it financially stable? What do students complain about? Which programs are strongest? Do graduates find jobs? Why should I go here instead of somewhere else?
Run those questions through several AI platforms and look for patterns.
You may find outdated information that needs correcting or great institutional stories that haven’t traveled far enough beyond your own website. You may discover that your strongest programs are practically invisible, or that the brand position leadership loves has very little external evidence supporting it.
You may also discover that AI has surfaced something everyone inside the institution already knows but nobody particularly wants to put on a PowerPoint slide.
That’s useful, too.
Because the point isn’t to teach the machine your talking points. It’s to understand the reputation your institution has accumulated when nobody in marketing gets to control the answer.
This is bigger than marketing
AI reputation belongs in the cabinet room because communications can strengthen institutional storytelling, build earned-media visibility, improve executive positioning, fix content gaps, and make credible evidence easier to discover. But communications cannot manufacture evidence that doesn’t exist.
Sometimes the right recommendation is a communications strategy. Sometimes the right recommendation is to fix the thing people are talking about.
Knowing the difference is the job.
For years, institutions have asked what people think about them. Then we started asking what Google knows about them.
AI introduces a more uncomfortable question:
When everything the world knows about your institution gets condensed into one answer, what story survives?
I’d want to know before the president asks.