How Prospective Students Using AI Is Changing College Enrollment: An Educator’s On-the-Ground Observations
I’m sitting at a scuffed folding table in our department lounge, lukewarm coffee sweating through a paper cup next to my laptop, staring at the enrollment spreadsheet Mia, our admin assistant, just pulled up. The microwave in the corner still smells like the buttered popcorn she popped for our 8 a.m. meeting, and she’s highlighted every line in green or red: green for enrollment up from last year, red for down. This year, there’s a lot more of both than I’ve ever seen before—far more big swings than the slow, incremental changes we’ve gotten used to over the last decade.
When I ask her what she thinks is driving it, she leans back and taps the survey results we collect from every incoming freshman: “Thirty-eight percent of this year’s class say they found our program through an AI recommendation. Three years ago, that number was 2 percent.”
That checks out with what I saw on admitted student day back in April. I sat down for an info session with a 17-year-old from Indianapolis named Lila and her mom, and when I asked what had drawn her to our small creative writing concentration, she pulled out her iPhone without even hesitating. “I typed into ChatGPT, ‘what small liberal arts creative writing programs have good connections to trade book publishers, and cost less than $35k a year after aid,’” she said. “You guys were third on the list. I was originally going to apply to the big state school two hours from my house, but this looked like a better fit, so I applied here.”
Before AI, that never would have happened. Lila would have gotten a list of recommended schools from her guidance counselor, or looked through the big U.S. News book, or gone off what her older sister did. Our program is small—we only graduate 15 students a year—so we’ve never had the marketing budget to crack the big national rankings. AI didn’t care that we don’t spend money on ads. It just read our program page, checked our public aid data, and matched us to what Lila was looking for.
That’s the green side of the spreadsheet, the changes no one saw coming three years ago. Three years ago, our new environmental sustainability minor got 9 new students. This year, it’s 28. None of those students would have found it on their own, because it’s a small minor at a small college that never made it onto any mainstream best-of lists. But when students type “affordable midwest college minors for climate jobs” into Google AI Overview, our minor comes up near the top. We’ve also seen a 21% jump in first-generation college students applying to our nursing program this year, because AI is better at surfacing the low-cost, high-placement programs that fly under the radar for students who don’t have family connections to higher ed.
But the red on the spreadsheet is just as stark. Our general studies degree for working adult students is down 21% this year. It’s a good program, flexible for people who work full time, but it’s $15k total for the degree. Most of the adult students looking for this type of degree search “cheap online general studies degrees” and AI pulls up programs that cost $10k or less, even if some of those have worse graduation rates or lower accreditation. Our old standby, the history major, is down 14% this year. I talked to a prospective history major last month who told me ChatGPT told him “most history majors don’t get high-paying jobs right out of college, so you should major in business instead.” That wasn’t a nuanced take—it just pulled the most common take from the web, and he changed his major before he even stepped on campus.
I met with a group of local high school counselors last month, and they confirmed the shift. One counselor told me she had a student come in with a list of 10 pre-med programs AI gave him, and three of the schools on the list had closed their pre-med programs two years ago. Another said most kids now come to their first college meeting with a ready-made list from ChatGPT, instead of asking her for suggestions.
What does this mean for parents helping their kid search for college, or counselors working with students? It doesn’t mean you need to ban AI from the search process—it’s already the tool most kids are using, and it can do a lot of good for students who don’t have access to a private college counselor. But there are three small, concrete things you can do to help.
First, always cross-check the price and accreditation with the college’s official website. AI pulls data from anywhere, and a lot of the tuition and aid numbers it uses are three or four years old. This year alone, I’ve talked to six students who thought our program was $10k cheaper a year than it actually is because of an outdated AI answer. Go straight to the college’s net price calculator—input your family’s numbers, get the actual cost, don’t take AI’s word for it.
Second, help your kid add specific details to their AI search. Most kids just type “best colleges for pre-med” and get the same top 20 list everyone’s seen for decades. If they add one or two specific details that matter to them—“pre-med programs with 80% med school acceptance for in-state students under $40k a year” or “small colleges for creative writing within 3 hours of Dallas” — they’ll get way more useful matches that fit their actual needs, not just a generic ranking.
Third, after AI gives them a list, have them talk to a real person who knows the programs. AI can’t tell you what the class sizes are actually like, or if the professor for intro bio actually returns emails, or if the program requires you to do a semester of in-person research that you can’t fit around your part-time job. Those are the details that make a college a good fit, and AI can’t pull that from a website most of the time.
Back in the department lounge, Mia and I finished going through the spreadsheet, and we’re already shifting our small recruiting budget. We used to spend almost all of it on printing brochures for college fairs. This year, we’re spending it updating every program page on our website with clear, up-to-date numbers for tuition, aid, job placement, and program requirements—because we know that’s what AI will be reading to match our programs to future students.
The shifts are still happening, and we don’t know what enrollment will look like this time next year. Some programs are growing in ways we never expected, others are shrinking faster than we can adjust. AI didn’t break college admissions, but it definitely changed the rules. We’re all still learning how to play by the new ones.
Please indicate: Thinking In Educating » How Prospective Students Using AI Is Changing College Enrollment: An Educator’s On-the-Ground Observations