For years, colleges were quiet about using AI in their own review process. That's changed. Several schools now say so outright, and the practice has moved from experimental to standard fast enough that it's worth understanding exactly what's happening to your file after you hit submit.
The Numbers Behind the Shift
A survey of admissions offices found that roughly eight in ten colleges planned to use AI somewhere in their admissions process for the 2024 cycle. Within that group, about seven in ten reported AI touching recommendation letters and transcripts, and six in ten said it was involved in reviewing personal essays in some capacity.
Not everyone in the field agrees on how far this goes. At least one vice provost publicly pushed back on the idea that admissions offices are using AI to "read" files in any meaningful sense — a reminder that practices vary widely by school, and not every institution is transparent about where the line sits.
Where AI Actually Shows Up
Based on what schools have disclosed, AI in admissions offices tends to cluster around a few specific jobs, not full decision-making:
First-pass sorting. Some schools use AI to check objective criteria — GPA thresholds, test score ranges, required coursework — before a file ever reaches a human reader. This is less "AI decides" and more "AI triages," similar to how resume screening software works in hiring.
Writing and grammar checks. UNC Chapel Hill has stated directly that its admissions office uses AI to do a basic evaluation of grammar and writing quality on essays. Importantly, the school has also said no admission decisions are made by AI — it's a supplementary check, not a verdict.
Applicant-facing chatbots. Separate from file review, many schools now run AI chat tools that answer routine questions about deadlines, financial aid, and program details around the clock, freeing staff time for more complex applicant questions.
Predictive enrollment modeling. Some institutions use AI on the back end to forecast yield — how likely an admitted student is to actually enroll — which can influence decisions on the margins, particularly for waitlist movement and merit aid.
What This Means for Your Application
If a school is using AI for a first-pass grammar and coherence check, that raises the floor on basic writing quality — typos and structural sloppiness are more likely to get flagged before a human ever sees the file. It's a good reason to proofread carefully regardless of how talented a writer you are.
If a school is using AI to screen for baseline eligibility, your numbers need to actually clear the stated ranges. This is less about gaming a system and more about accuracy: don't assume a borderline GPA will get context and nuance from a human reader if it doesn't survive an automated first pass.
None of this replaces a holistic read by an actual admissions officer at any school that still promises one — and most selective schools still do. AI is filling in around the edges of a process that remains, for now, fundamentally human where it counts most: the final decision.
The Honest Uncertainty Here
Admissions offices don't uniformly disclose their internal tools, and practices differ enormously between a large public university processing 100,000 applications and a small liberal arts college reading 8,000. Treat any specific claim about "how School X uses AI" with some skepticism unless it comes directly from that school's admissions page — and when in doubt, a quick email to the admissions office asking directly is a completely normal thing to do.