Almost every student I talk to asks some version of this. 'Will they know I used AI?' The honest answer: admissions officers aren't running your essay through a detector. They're reading it. And they've been reading thousands of essays a year long enough to know when something doesn't sound like a 17-year-old wrote it.
What actually gives it away
It's not one obvious tell. It's a pattern.
None of this requires software. It requires an experienced reader. Every admissions office has one.
The schools that are explicit about it
Policies are changing fast, so check each school's instructions rather than assuming. Some schools — the UC system, for instance — explicitly disqualify applicants who submit AI-generated content. Others draw a harder line at drafting than at brainstorming help. Don't assume the policy is uniform across your list.
What's actually safe
The question isn't AI or no AI. It's whether the tool is writing for you or helping you write.
What's risky is anything that hands you finished sentences. If a tool writes a paragraph and you paste it in with light edits, you've outsourced the exact thing admissions officers are reading for.
The bigger problem isn't detection
Even setting aside the policy risk, there's a more basic issue: AI-written essays are technically polished and emotionally empty. They don't get students into selective schools whether or not anyone flags them. The self-reflection required to write a strong essay isn't an obstacle to skip — it's the entire point of the exercise.
Every word in your final draft should be one you'd stand behind in an interview. Because at plenty of schools, you might have to.