Thousands of students will get their college decisions faster this year because of artificial intelligence being used by admissions departments to read essays and review transcripts.
The use of AI aims to streamline work for staff and cut costs for colleges facing funding pressures, with schools like Virginia Tech and the California Institute of Technology using the technology to accelerate the review process.
Some AI experts worry that reliance on the technology could introduce bias into the decision-making process, with concerns that it may favor a certain type of student and leave disadvantaged groups behind.

Thousands of students will get their college decisions faster this year because of artificial intelligence.

Admissions departments across the country are turning to AI for help reading essays and reviewing transcripts, aiming to streamline work for their staff as higher education faces funding pressures. With corporate America using the technology to screen job applicants, schools like Virginia Tech and the California Institute of Technology are taking a page out of hiring managers’ playbook.

It’s the latest indication of how large language models from the likes of OpenAI Inc are drastically altering US education. College professors are integrating AI into their teaching methods, and students are turning to ChatGPT for help with essays, despite widespread warnings of unaddressed risks.

For students eager to hear back on admissions decisions, typically released in April, a faster decision process may reduce a stressful waiting period. But some AI experts worry that reliance on the technology could introduce bias into the decision-making process, already a fraught issue as President Donald Trump’s administration cracks down on diversity, equity and inclusion practices.

“I don't know any parents who are excited about AI being used to evaluate one of the most important decisions in their kids’ life,” said Meredith Broussard, a professor at New York University whose research focuses on AI ethics. “Using AI to evaluate essays rewards predictability, it rewards people who follow the exact mathematical format that a particular AI is looking for and that's actually the opposite of what we want in student writing.”

Streamlined Process
For colleges contending with federal funding freezes and the prospect of lower enrollment, using AI could help cut costs by reducing the number of humans needed to read applications, said Terry Crawford, co-founder of InitialView, a company used by many colleges to allow student videos as part of applications.

Consider the application process at Virginia Tech. In a typical year, students would answer four short-essay questions, which would then be reviewed and scored by two human readers. A third would be called on if the first two scores varied greatly. For last year’s record 58,000 applications, there were about 500,000 essays to read.

This year, both a single reader and AI will give scores for each essay question, and if there’s a discrepancy, an additional human reader will also give a score. That’s so far saved an estimated 8,000 hours, according to Juan Espinoza, vice provost of enrollment management. The school is planning to release admissions decisions about one month earlier than usual, thanks to AI.

“Humans still have a major role,” Espinoza said. “But if there's opportunities and subsections of the application where it can help accelerate the review or summarize in some way, I think those are opportunities that colleges should consider and investigate.”

Georgia Tech is also using AI to review transfer student transcripts in order to get decisions out quicker, said Rick Clark, executive director of strategic student access at the school. The school is currently developing a tool to analyze historical and census data to estimate how much a student would likely receive in financial aid, which is often difficult for admitted students to discern.

Other colleges are essentially using AI to guard against applicants relying too heavily on tools like ChatGPT in their applications. At the California Institute of Technology, students who submit research as part of their application can opt to respond to a series of AI-generated questions through a video platform called Viva. Created by InitialView, the tool will review the research papers and generate a series of questions, which the students will answer in a recorded video.

Risks for Applicants
While AI could improve efficiency for admissions departments, the danger would be if colleges rely on it too heavily to make decisions or rankings, said Daron Acemoglu, an economics professor at the Massachusetts Institute of Technology.

If multiple universities use the same AI models, they might start making the same decisions that favor a certain type of student, he added. Allowing students to use AI for materials in their applications comes with similar risks.

“Some students will be able to use more advanced AI models, and these may be students who are better off financially or are better connected,” Acemoglu said. “In the meantime, some disadvantaged groups will be left further behind.”

These limitations are part of the reason why some schools, including Brown University, have cracked down on students using AI for their applications, or clearly laid out appropriate use cases. When applying to Brown, an applicant can use AI to assist with spelling and grammar review, but the content of all short-answer questions and any other material submitted must be the work of that student.

Cornell University permits students to use AI for brainstorming essay topics and reviewing grammar, while unethical uses include using AI to write essays, translating an essay written in another language or creating images for a required portfolio.