As top students are starting their college applications this summer, parents are asking me: Should my teenager use AI to write their college essay?
As a college admissions expert, I tell most parents that the honest answer depends entirely on how and when AI is used.
For starters, a student producing a lazy, generic essay is nothing new. Families have long been able to buy one from an essay mill, copy a “winning” template from a database, or pay a consultant to write it. AI did not invent the shortcut; it just made it cheaper and universal.
A recent study of more than 370,000 application essays, featured in the New York Times, found that after ChatGPT arrived, college essays became more sophisticated and more colorful, while sounding more similar to one another. They read better and said less.
So if your kid decides to use AI in writing their admissions essay, here’s how they should and when they should.
What makes a great college essay
The first thing I teach parents is that a student is not writing for an institution. They are writing for a person: an admissions officer who reads hundreds of essays every week and is exhausted.
That person is not choosing the most impressive applicant. They are choosing the most memorable one.
A great essay is not a list of accomplishments, because the rest of the application already highlights those. Instead, it earns attention in three ways:
- The essay says something genuinely insightful — an idea the reader has not considered that way.
- The essay shows real reflection and self-awareness — a student who has examined their own life rather than just lived it.
- The essay makes the reader think: “I would love to have this person as a roommate or a classmate, someone I want to know.”
How my students get into Ivy League schools
When I coach students, we don’t start with writing. We start with moments. Most students want to write about a category: three years of violin, a season of debate, a summer of volunteering. Categories are forgettable because everyone has them.
Before anyone writes a word, I have students spend a week mining their own lives for three kinds of material.
The first is their mess-ups or the failures that make a person relatable. The second is their sacrifices: what a student gives up, time or comfort, because that reveals what they actually value. The third, and the hardest, is their contradictions: the atheist who studies holy texts, the football star who is fearless on the field and afraid of contact off it. Those surprises are what make a reader lean in.
Only after that week of brainstorming do my students write a first draft, and it is usually rough. The ones who get into the most competitive schools do not stop there. They write 10, 15, sometimes nearly 20 complete drafts, each one closer to the truth.
The thinking is the work, and the thinking is the part no one can hand off.
Where AI could fit, and where it should never go
AI won’t ever have your child’s insight, memory, or voice. What it could do, after a student has built an honest draft entirely on their own, is push that draft to be sharper.
Used as a critic rather than a writer, AI can challenge a student to be more vivid and precise.
Here are a couple of prompts I would trust a student to use, and only on a draft that is already fully theirs:
Does this part sound generic, or am I making assumptions I have not examined yet?
Is anything here confusing or underdeveloped?
Those prompts use AI as a sparring partner, not a ghostwriter. It can only help because the ideas, the moments, and the voice were there first.
The AI panic is real, but the problem underneath it is old, and so is the solution: a real student saying something true in a voice that is unmistakably their own. No chatbot has ever done that.
Help your teenager do the thinking first, and if AI comes in at all, let it come in only at the end, to sharpen an authentic essay rather than replace one.
Steve Gardner teaches Leadership and Impact at Harvard Summer School and is the founder of The Ivy League Challenge.
Read the full article here
