Should My Kid Use ChatGPT for the College Essay?

Should My Kid Use ChatGPT for the College Essay?

Two uses at most: asking brainstorming questions, and checking grammar and spelling on a finished draft. Those are the only uses that Caltech and QuestBridge both call ethical. Brown permits only the spelling-and-grammar part. Anything that writes, outlines or translates the essay is outside those rules.

So do not ban AI, and do not police it. Agree on those two uses in writing. Then judge the draft by one test from the Coalition for College's own guidance: can your child explain every choice in it? This page gives you the rules side by side, a family agreement you can copy, and a way to raise the question without an accusation.

Every college and platform page quoted below was checked on October 30, 2026. AI rules change each cycle, so open each college's own page before your child submits.

On this page: What your child signs · What colleges allow · A family agreement · How to tell without accusing · If the draft is already AI-written · Questions

What your child signs

Your child, not you, certifies the application. The Common App's fraud policy lists this among its examples of fraud:

"Submitting plagiarized essays or other written or oral material, or intentionally misrepresenting as one's own original work: (1) another person's thoughts, language, ideas, expressions, or experiences or (2) the substantive content or output of an artificial intelligence platform, technology, or algorithm"

Clause (2) is the AI clause. Clause (1) is the one that covers parents. Our guide to how much parents should help with the college essay quotes the full certification and maps parent actions to it.

Some colleges add their own rule on top. Caltech's Fall 2027 guidelines say: "Failure to comply with the Ethical Use of AI guidelines may result in the rescission of your admission to Caltech."

What colleges and platforms allow

Four published positions, in their own words:

The Coalition page is advice written for students and families. It is not a college's rule. It also points out that "some colleges have specific AI policies, and admissions offices follow those guidelines when reviewing applications."

Where the rules disagree

• Outlines. The Coalition guidance lists "Creating an outline" as appropriate. Caltech and QuestBridge both list AI outlining as unethical.
• Revising sentences. The Coalition guidance allows it "as long as the ideas are still yours." Caltech and QuestBridge allow only a grammar and spelling review of a completed essay. Brown allows only "spelling and grammar review".
• Brainstorming. Caltech and QuestBridge allow AI to generate questions. Brown's page does not list brainstorming as a permitted use.
• Translation. Caltech and QuestBridge both bar translating an essay written in another language. This matters for multilingual families: drafting at home in your language and translating with AI is on both lists of unethical uses.

Our recommendation: write your family rule to the strictest college on your child's list. The basis is the table: the uses that every source above allows, except Brown, are brainstorming questions and a grammar check on a finished essay. If Brown is on the list, drop brainstorming too. One rule for the whole list is easier to keep than a different rule per college.

The countercase: if your child's list has no college with its own AI rule, the Coalition guidance describes a looser approach, and some families will follow it. That is a real choice, but the Common App clause still applies to every college on the list. If a school publishes no AI rule, the certification inside the application is what you sign, so read it there. Our AI policies directory lists published college rules.

What is unknown: none of the four pages explains how a college checks for AI use.

A family agreement you can copy

Constructed example: a one-page family AI agreement

Constructed example, written by us, not a real family's agreement. Edit it to fit your child's college list.

Our college essay AI agreement

1. AI may ask questions to help brainstorm topics. It does not suggest topics or write sentences. (If Brown is on the list, no brainstorming with AI.)
2. After a draft is finished, AI may check spelling and grammar. The writer accepts or rejects each change.
3. No AI outlines, drafts, rewrites or translations. No pasting AI text into the essay.
4. All notes and drafts stay in one document with version history turned on.
5. A parent reads each draft once and asks questions. A parent does not write or reword sentences.
6. If either of us is unsure about a use, we read that college's own AI page before doing it.
7. Before submitting, the writer can answer four questions out loud: why this topic, what the story means, why this structure, and why these words and details.

Signed: __________ (student) __________ (parent) Date: ______

Why it is built this way:

• Items 1 to 3 are the overlap in the table above.
• Item 4 gives your child a record of their own process, draft by draft.
• Item 5 applies the same rule to you that the fraud clause applies to everyone: another person's "language" and "ideas" do not belong in the essay.
• Item 7 is the Coalition's own self-test, turned into a habit.

Caltech gives a simple test for any use not on the list: "ask yourself whether it would be ethical to have a trusted adult perform the same task you are asking of ChatGPT. Would a teacher be able to review your essay for grammatical and spelling errors? Of course! Would that same teacher write a draft of an essay for you to tweak and then submit? Definitely not."

How to tell without accusing

You may already suspect a draft. Starting with "Did you use AI?" puts your child on the defense and tells you little. Start with the essay instead.

The Coalition guidance gives the test. It asks whether a student can "comfortably talk about" four things: "Why you chose the topic", "What the story or example means to you", "Why you structured your essay the way you did" and "Why certain words or details felt important". Its conclusion: "If the answer is no, you may be relying too heavily on AI."

Constructed example: a five-minute conversation

Constructed example, written by us, not a real conversation.

Parent: I read it. Why did you pick the summer job over the robotics team?

Student: Because the robotics one sounded like everyone else's.

Parent: Your last paragraph says you "learned the true meaning of responsibility." What does that mean for you, specifically?

Student: I don't know. It sounded like a good ending.

Parent: What actually changed after that summer?

Student: I stopped asking my manager what to do when the register was off. I just counted the drawer.

Parent: That sounds like a better ending than the one you have.

What happened there: the student could explain the topic choice (question 1). They could not explain the last line (question 4). The parent did not accuse, rewrite or suggest words. The new ending idea came from the student. Whether a chatbot wrote the old ending or a habit did, the fix is the same.

The Coalition page makes the same point: "Essays that sound generic or disconnected from the rest of an application often raise concerns, regardless of how they were produced."

Why a detector score will not settle it

Our view: do not use an AI detector score to settle the question at home. A score is not proof, and a wrong one starts a fight you cannot resolve.

The basis is published research. A 2023 study of AI-text detectors found that the detectors it tested "consistently misclassify non-native English writing samples as AI-generated, whereas native writing samples are accurately identified." So a score alone can point at an honest essay by a child who learned English as a second language. The four questions above tell you more. Our guide to whether AI detectors are accurate on short essays covers short text.

The countercase: if a teacher or counselor has already raised a detector result, the conversation has started without you. Our guide for students whose essay was flagged as AI but they wrote it covers what to do next. Counselors and teachers have their own guide to telling whether a student essay is AI.

If the draft is already AI-written

If the conversation shows that a chatbot wrote the draft, do not ask your child to "edit it until it sounds like them." Editing leaves the AI's content in the essay, and the Common App clause covers "the substantive content or output" of AI, not only its exact words. Start again from what really happened:

1. Close the AI draft. Do not keep it open in another window.
2. List the real events from memory: what happened, in order, with names, places and numbers.
3. Pick one event from the list that answers the prompt.
4. Write a new draft from the list. Short and plain is fine.
5. Run the four questions again on the new draft.

Our guide to making a college essay not sound like AI covers the rewrite for students. For the student's side of the rules, see the student guide to using AI on college essays.

Questions parents ask about AI and the college essay

Can my child use AI to brainstorm?

For questions, at Caltech and QuestBridge, yes: both allow AI to generate brainstorming questions. Brown's page permits only "spelling and grammar review", so if Brown is on the list, no.

Does a grammar checker count as AI use?

At Brown, a spelling and grammar review is allowed "in the same way as any other platform that supports basic proofreading." Caltech and QuestBridge also allow a grammar review of a completed essay.

We speak another language at home. Can my child draft in it and translate?

Not with AI at Caltech or QuestBridge. Both list translating an essay written in another language as unethical. Our guide to writing a college essay in your native language, then translating covers the rules and a bilingual drafting method.

Can I read the essay at all?

Yes. Reading and asking questions is not the problem. Writing or rewording is. See how to give feedback on your teen's college essay for a method that stays on the right side of the line.

Get a free review of your Common App essay

Get a free Quick Review of your Common App essay against GradPilot's narrative craft rubric. Your child pastes their own draft and reads the feedback, so the writing stays theirs. It reads whether the essay builds on one main moment, uses details that change the story, and ends in a way that belongs to it. The Common App essay review page explains how it works.

More for parents of applicants:

• How much should parents help with the college essay?: the allowed and not-allowed table.
• Can my parents see my Common App essay?: who can open your child's account and essay.
• All Common App essay guides: the full hub.

Sources

Retrieved October 30, 2026.

• Common App, Fraud Policy: the fraud clause, including the AI clause.
• Coalition for College, AI and the College Essay (My Coalition Counselor; the page shows "Feb 18" with no year): appropriate uses, uses to avoid, the four-question self-test, "regardless of how they were produced".
• Caltech, Ethical Use of AI: Guidelines for Fall 2027 Applicants: ethical and unethical uses, the trusted-adult test, rescission.
• QuestBridge, What is QuestBridge's policy toward artificial intelligence use in the Writing section? (updated October 16, 2026; read through the help center's public article feed): ethical and unethical uses.
• Brown University, Integrity in the Application Process: the AI rule and the spelling-and-grammar exception.
• Liang and co-authors, arXiv 2304.02819 (2023): detector misclassification of non-native English writing.