This shift is driven by a profound generational change. While the Gen Z students currently on campus grew up with smartphones and social media, the incoming waves of prospective students, members of Gen Alpha, are digital natives of a different order: a generation raised on screens from infancy and coming of age in a world where AI is seamlessly built-in.
Because these students have lived alongside advanced technology for most of their lives, they naturally turn to it when making one of the biggest decisions of their adulthood. So it makes sense, then, that fewer students are reaching out to prospective institutions directly to learn more. Instead, they’re using large language models like ChatGPT to help choose colleges, determine majors, and learn which careers have more earning power.
A February 2026 Education Advisory Board survey found 46 percent of students use AI during their college search, up from 26 percent a year ago. And 18 percent say they’ve removed an institution from their list because of AI. In short, AI is helping them design their futures and lives.
Adapting the Digital Front Door
In response, many college administrators are working to get on board by equipping website chatbots with accurate, timely answers because students rely heavily on them for decision making. Algorithms expose important details of the what, where and why in student searches. The findings give college recruiters real-time insight into how to approach and meet their needs.
At Rollins College, a small liberal arts institution in Winter Park, Fla., vice president of enrollment and marketing Faye Tydlaska says AI has pushed the institution’s leaders to rethink the structure of their website and “how we put information on the web.”
“We want AI and the new tools of emerging technology to help enhance what we’re able to do and really to be extensions of our office,” she said. “We can’t be everywhere. We’re small but also making sure that humans are always tuned in. We never want to turn over all our communications to AI.”
Balancing of Accuracy and Trust
Tydlaska, who has been in her role for ten years, says creating information for chatbots is a delicate dance between accuracy and effectiveness. Trust in college marketing comes to life through attention to the little details, she continues, adding that when admissions officials do get an opportunity to speak to students or parents, they’re sure to make the most of their time and answer questions that chatbots may not.
“When our counselors are able to speak to students, they make sure they’re having really substantive conversations with them, because we don’t want students and parents to get wrong information from a chatbot,” she explains. “Actually, it’s a very heavy lift for our marketing team to ensure that AI responses from our website are accurate, the tone is correct and we’re conveying what we want to convey.”
Avoiding AI hallucinations — language and information generated by artificial intelligence models that is incomplete or untrue — is another reason for human interaction. Ensuring that students and parents are receiving correct information regarding programming, enrollment, or aid, is one of the best ways to build trust and confidence in applicants.
Robert Clougherty, AI Strategy and Innovation Lead at the Alliance for Innovation & Transformation, says student and parent trust in the institution is mainly affected by whether colleges can stand behind their public messaging. “The deciding factor isn’t whether you use AI, it’s whether you can answer for what it does,” Clougherty says. ”Trust in marketing has never come from the channel; AI doesn’t change that rule; it raises the stakes on it.”
“A system that personalizes outreach a student can’t see the logic behind, or that makes a promise the institution can’t keep, erodes trust faster than any brochure ever could and at scale,” continues the former college CIO, dean, and provost.
The Small School Advantage
Clougherty also believes smaller institutions have a particular advantage that larger ones don’t when incorporating AI in their process.
“The assumption everyone arrives with is that wealthy institutions will build sophisticated AI recruitment while smaller ones get left behind because they can’t afford it. But the tools are cheap and getting cheaper,” he says.
“What’s genuinely scarce is the capacity to govern them well, and on that front, the small institution has a structural edge the flagship doesn’t: fewer layers, fewer systems, a smaller team that can see the whole funnel and change course in a semester instead of a planning cycle. The large university has more resources and more inertia, and in a pivot this fast, inertia is the liability. The catch is that the advantage is potential, not kinetic,” Cloughetry adds.
Streamlining Admissions
Admissions officers are not just using AI to help with recruitment — some are also using it to help decide which students should be admitted on campus as well.
In the Winter 2026 edition of National Association for College Admission Counseling’s Journal of College Admission, Matt Lopez, deputy vice president for academic enterprise enrollment, at Arizona State University detailed how ASU holistically embraces AI to eliminate time-intensive back-end tasks. By automating transcript processing and speeding up admissions decisions, the technology allows the university to upskill its staff quickly and prioritize more meaningful human interaction.
But despite those advantages, using AI in admissions decisions evokes even deeper questions about ethical use of technology on campus. One growing concern is whether AI will take away from campus diversity, given the known biases in algorithmic systems.
‘Redlining by Proxy’
As with any computing technology, the algorithm’s choices are made according to patterns. If an institution is already lacking in diversity, it is likely the algorithm's idea of who belongs on campus will be shaped by who is already there.
If enrollment numbers show a decrease in people of color, who — or what — is to blame?
Clougherty says that AI will only follow the patterns of how schools already approach diversity. This will eventually reveal the institution’s truth in managing enrollment diversity practices.
“A college can end up quietly recruiting toward the students who resemble the ones it already enrolls, narrowing access while its dashboard reports success,” he adds. “That’s the failure mode: not the system being wrong, but the system being confidently adequate while it narrows the door.”
“The specific risk is what I’d call redlining by proxy,” he continues. “Enrollment AI rarely uses protected characteristics directly; it doesn’t need to. It uses zip code, high school, distance from campus, engagement signals. All the variables correlate tightly with race and income. The model never ‘sees’ the protected class. It rebuilds the same pattern through the back door, laundered through data that looks perfectly neutral, and presents it back as a yield optimization.”
Tydlaska says she has noticed how other colleges that include AI in their student acceptances or denial are working to level the playing field and remove biases based on financial information.
“I’ve heard a lot of my colleagues who are experimenting with AI say it’s really important for them to understand the biases that AI can bring,” she says. “Particularly in a financial aid scenario, and to be able to train their staff to identify that so they can go ahead and train AI not to move those forward. But it’s certainly a huge concern.”
