Admissions directors say AI tools are being incorporated into the way student applications are reviewed and analyzed. This can be a sensitive topic, and not all universities are willing to talk about it, but higher education is one of the many industries where artificial intelligence is rapidly taking over tasks previously handled by humans.
In some cases, experts say, schools are quietly incorporating AI into their evaluation processes. Some companies are touting the technology's potential to speed up application reviews, reduce processing times, and even perform some tasks better than humans.
“Humans get tired. Some good days are better than others. AI doesn't get tired. It doesn't get grumpy. It doesn't have bad days. AI is consistent,” said Juan Espinoza, Virginia Tech's vice provost for enrollment management.
Virginia Tech plans to debut an AI-powered essay reader. The university hopes the tool will help it sort through tens of thousands of applications, allowing it to notify students of admission decisions a month earlier than usual.
The university emphasizes that it does not rely on AI to make admissions decisions and primarily uses AI to verify transcripts and eliminate data entry tasks. But artificial intelligence also plays a role in student assessment. A select number of schools are deploying AI tools to scrutinize the increasingly curated application packages that some students develop with the help of high-priced admissions consultants.
Caltech will introduce an AI tool this fall to examine the “authenticity” of students who submit research projects with their applications, director of admissions Ashley Parry said. Students upload their research to an AI chatbot, where they are interviewed about it on video and reviewed by Caltech faculty.
“This is a measure of credibility: Can you make an intellectual case for this research? Is there a certain joy in your project? That passion is important to us,” Parry said.
Ruby Bhattacharyya, chair of the National College Admissions Counseling Association's Admissions Practice Committee, said the use of AI is such a new trend that it's difficult to measure its adoption. NACAC updated its ethics guide this fall, adding a section on artificial intelligence. The document asks universities to ensure their use is “in line with our shared values of transparency, honesty, fairness and respect for the dignity of students.”
Some schools face backlash over use of AI
The University of North Carolina at Chapel Hill faced a barrage of negative feedback from applicants, parents, and students after the student newspaper, the Daily Tar Heel, reported in January that the school was using AI to evaluate the grammar and writing style of applicants' essays.
The university declined to comment for this article, pointing to its admissions website, which had been updated in response to criticism. “UNC uses an AI program to provide data points on students’ common application essays and school transcripts,” the website states. All applications are “comprehensively evaluated by extensively trained human application evaluators.”
Virginia Tech's Espinoza said he has been contacted by several universities interested in the new technology but wary of the backlash. “The feedback from a lot of my colleagues is, 'You roll this out, we're watching you, and let's see how everyone reacts,'” he says.
He stressed that the AI reader, which the school spent three years developing, is only used to check the scores of human readers' essays.
Until this fall, each of the four short answer essays submitted by Virginia Tech applicants was read and graded by two people. Under the new system, one of those leaders will be an AI model trained on past applicants' essays and scoring rubrics, Espinoza said.
If the AI and human reader's opinions differ by more than 2 points on a 12-point scale, a second person intervenes.
Like many universities, Virginia Tech has seen a significant increase in the number of applicants since making the SAT optional. Last year, a record 57,622 people applied for the 7,000-person freshman class. Despite having 200 essay readers, schools are struggling to keep up and end up notifying students later.
This AI tool can scan approximately 250,000 essays in less than an hour, compared to an average of 2 minutes per essay for human readers. Based on last year's application pool, “we're saving at least 8,000 hours,” Espinoza said.
Universities recognize the benefits of AI tools for applicants
This message is a sensitive one for universities, many of which now require students to certify that they are not using AI unethically in essays or other parts of their applications. But schools argue that AI tools can help admissions offices eliminate errors in tasks such as uploading transcripts, simplifying the process for students.
Georgia Tech will introduce an AI tool this fall to review transfer students' college transcripts, replacing the need for staff to manually enter each course into a database. Richard Clark, the school's executive director of enrollment management, said this will allow the school to more quickly notify applicants of the number of transfer credits they will receive, reducing uncertainty and wait times.
“There's going to be a lot more delays and stress and inevitable errors. AI is going to beat that. I'm really excited,” Clark said. The school hopes to expand the service to all high school transcripts soon. Georgia Tech is also testing AI tools for other uses, including one to identify low-income students who may be eligible for federal Pell grants but may not be aware of it.
Stony Brook University in New York is also testing AI tools for a variety of tasks, including using artificial intelligence to review applicants' transcripts and summarizing student essays and letters of recommendation to highlight considerations for admissions officers, said Richard Beatty, the school's senior vice president for admissions management.
“Maybe the student was battling an illness during their sophomore year. Or maybe a parent has passed away or they're at home caring for a sibling. All of those things are important and it allows the counselor to look at the transcript in a different way,” Beatty said.
Emily Pacheco, founder of NACAC's AI and Admissions Special Interest Group, said universities are interested in AI summaries of transcripts, extracurricular activities and letters of recommendation that tell a student's story in a more understandable way to human readers.
“Humans and AI working together is the key right now. We can significantly improve every step of the process, from reading transcripts to reviewing essays to telling us things we might have missed about a student,” said Pacheco, former assistant director of admissions at Loyola University Chicago. “Ten years from now, all bets are off. AI will probably get students admitted.”
