Should You Tell Colleges You Used AI on Your Application?

You used ChatGPT to brainstorm essay topics. Or Grammarly to polish your personal statement. Maybe you asked Claude to give feedback on a draft. Now you're staring at the Common App wondering: do I need to tell anyone?

This is the question hundreds of thousands of applicants are Googling right now. And the answer is more nuanced than most advice sites suggest. We analyzed the AI disclosure policies of 150+ universities using our L/D/E rubric to give you a concrete, data-driven framework for deciding.

The Numbers: Most Schools Don't Even Ask

Here is the disclosure landscape across the 150+ schools in our database:

The headline finding: more than two-thirds of colleges have no disclosure mechanism at all. There is no box to check, no supplemental question, no place in the application where you are specifically asked about AI.

But that does not mean disclosure is irrelevant. Let's break down what each level means for you.

D0 Schools (68%): No Disclosure Mechanism

Schools like Harvard (D0/L0/E1), MIT (D0/L0/E0), Yale (D0/L3/E3), Princeton (D0/L1/E1), Duke (D0/L2/E1), Northwestern (D0/L0/E0), and Cornell (D0/L2/E1) have no formal place in the application for AI disclosure.

What this means for you: There is no question to answer and no checkbox to tick. These schools are not actively soliciting information about your AI use through their application forms.

But here's the catch: The Common Application's fraud policy still applies. When you submit through the Common App, you certify that your work is your own. Yale goes further, explicitly stating that "submitting the substantive content or output of an artificial intelligence platform constitutes application fraud" with consequences including admission revocation or expulsion. So the absence of a disclosure box is not a green light to submit AI-generated content.

Practical advice: If you only used AI for brainstorming or grammar checking at a D0 school, you do not need to go out of your way to disclose. There is no mechanism for it, and unsolicited disclosure could create confusion where none exists. Focus instead on ensuring your essays genuinely reflect your voice.

D1 Schools (7%): Voluntary Disclosure Encouraged

Schools like Stanford (D1/L0/E1), UPenn (D1/L1/E1), UVA (D1/L2/E2), Tufts (D1/L1/E2), Boston College (D1/L2/E2), and UC Santa Barbara (D1/L1) encourage transparency about AI use but do not require it through a mandatory field.

Stanford's approach is instructive. Their admissions page emphasizes that "essays should allow the applicant's genuine voice to come through," and their AI Advisory Committee has recommended "not using generative AI in admissions decisions or application review without careful assessment." The D1 classification means they want to know, but they are leaving the door open for you to decide.

UPenn's institutional AI guidance is direct: "Be transparent about the use of AI. Disclose when a work product was created wholly or partially using an AI tool."

What this means for you: If you used AI in a substantive way, like getting feedback on a draft or using translation assistance, disclosing proactively is a smart move at D1 schools. It demonstrates integrity and aligns with the institution's values.

Practical advice: Use the Additional Information section of the Common App or the school's supplemental materials to add a brief, confident disclosure. Keep it factual. Our AI Disclosure Generator can help you draft one in about 60 seconds.

D2 Schools (22%): The Application Asks You Directly

This is where disclosure shifts from optional to expected. Schools like Caltech (D2/L2/E1), UC Berkeley (D2/L1/E2), Swarthmore (D2/L1/E2), Auburn (D2/L2), Michigan State (D2/L2), and UNC Chapel Hill (D2/L0/E0) include a disclosure question or prompt in their application.

Caltech's approach is one of the clearest. Their application directly asks: "Did you receive any AI generated assistance in the preparation of your application materials?" They also specify what is acceptable (grammar checking, brainstorming) and what is not (drafting, outlining, translating). Critically, Caltech states that the disclosure is confidential and will not impact admissions decisions - they are trying to gather data, not punish honesty.

Swarthmore requires students to "acknowledge AI assistance and cite AI-generated content if used." Their policy explicitly prohibits using AI to outline, draft, translate, or rewrite significant portions of essays.

UC Berkeley follows the UC system-wide Statement of Integrity, requiring that "content and final written text must be their own" while allowing advice on editing and readability.

What this means for you: Answer honestly. These schools are asking a direct question, and lying or omitting information on a college application is far worse than admitting you used Grammarly for grammar checking. If the school says the disclosure will not affect your admissions decision (as Caltech does), take them at their word.

Practical advice: Be specific about what you used and how. "I used ChatGPT to brainstorm three potential essay topics, then wrote all drafts myself" is a strong answer. "I used AI" with no detail is not helpful to anyone.

D3 Schools (2.4%): Mandatory Attestation Required

Four schools in our database require you to sign a formal statement about AI use: Georgetown (D3/L4/E3), BYU (D3/L4/E2), NC State (D3/L3/E3), and SMU (D3/L4/E3, graduate programs).

Georgetown's undergraduate application states bluntly: "use of AI tools to complete any portion of the application is prohibited." This is not a disclosure prompt, it is a prohibition with an attestation requirement.

BYU is equally explicit: "You may not use generative AI tools (like ChatGPT) as you compose your responses." They back this up with the warning that "BYU may rescind the admission offer."

NC State requires applicants to certify that "My application materials were not created by another person or by a generative artificial intelligence system," while still allowing AI for brainstorming and structuring.

What this means for you: At D3 schools, there is no gray area. You are signing a statement. If you used AI tools for anything beyond what the school explicitly permits, do not apply without addressing it honestly or reconsidering your approach.

Practical advice: Read the exact wording of the attestation carefully. NC State, for example, allows brainstorming with AI but prohibits submitting AI-generated language. That is a meaningful distinction.

What Admissions Officers Actually Think

The data from your school's policy is only half the picture. The other half is what the humans reading your application believe.

A 2025 Kaplan survey found that 50% of admissions officers view AI use in applications unfavorably. Only 14% view it favorably. The remaining 36% are neutral or undecided.

This is a striking asymmetry. Even at schools where AI is technically permitted for brainstorming or grammar checking, half the people evaluating your application have a negative gut reaction to AI in the admissions process.

A foundry10 research study confirmed that AI is fundamentally reshaping how students approach applications, but the institutional response remains cautious and often skeptical.

The Authenticity Problem: AI Won't Sound Like You

A 2025 Cornell study found that AI-generated college essays systematically lack the markers of authenticity that admissions officers value. The study's conclusion is captured in its title: "AI can write your college essay. It won't sound like you."

Cornell's research showed that AI essays tend to be technically proficient but emotionally flat. They hit the structural notes of a good essay without the specificity, vulnerability, and quirks that make a personal statement personal. Admissions officers, who read thousands of essays each cycle, are increasingly attuned to this pattern.

This finding aligns with what schools across the spectrum are saying. Princeton's Dean of Admission noted: "I guarantee that any essay one writes with the help of AI is not going to be nearly as good or authentic." Duke has gone so far as to stop numerically scoring essays entirely, evaluating them holistically for content rather than writing quality, in part because of AI concerns.

The Equity Angle

One complication worth acknowledging: AI tools can genuinely help students who lack access to traditional essay coaching. First-generation students, students from under-resourced schools, and international applicants may benefit from AI assistance in ways that level a playing field that was never level to begin with.

But the equity argument cuts both ways. Students from wealthier backgrounds often have access to more sophisticated AI tools, premium subscriptions, and the digital literacy to use them more effectively. The student using free ChatGPT for brainstorming and the student using a $200/month AI writing suite with custom prompts are both "using AI," but the outcomes are very different.

This is why schools like Caltech and Swarthmore are trying to create disclosure frameworks that distinguish between types of AI use rather than issuing blanket bans. The policy landscape is evolving toward nuance, not prohibition.

Your Decision Framework

Based on our analysis of 150+ school policies, here is how to decide:

Step 1: Check your school's D code. Use our AI Policies directory to look up every school on your list. The D code tells you whether disclosure is expected.

Step 2: Categorize your AI use.

• Brainstorming and research (used AI to generate topic ideas, learn about programs): Low-risk, minimal disclosure needed
• Grammar and mechanics (Grammarly, spell-check, basic editing): Low-risk, generally accepted everywhere
• Feedback and revision (asked AI to critique a draft, suggest improvements): Medium-risk, disclose at D1+ schools
• Translation assistance (used AI to translate writing to English): Medium-risk, context-dependent, check school policy
• Content generation (had AI write sentences, paragraphs, or outlines you kept): High-risk, disclose everywhere it is asked, reconsider at D3 schools

Step 3: Match your use to the school's policy.

• D0 school + low-risk use: No disclosure needed. Focus on authenticity.
• D1 school + any substantive use: Disclose proactively. It shows integrity.
• D2 school: Answer their question honestly and specifically.
• D3 school: Read the attestation. Comply fully or reconsider.

Step 4: When in doubt, disclose. A brief, confident statement about using AI for brainstorming will never hurt you. An undisclosed use that gets discovered will.

Generate Your Disclosure Statement

If you decide to disclose, our AI Disclosure Generator creates a professional disclosure statement in about 60 seconds. It walks you through selecting the tools you used, categorizing your usage type, and produces a statement you can paste into the Common App Additional Information section, a Coalition App response, or a school-specific portal.

The tool is consistent with the framework in this article: it focuses on being specific, confident, and transparent without over-disclosing or being defensive.

Check Any School's Policy

Not sure about your school's disclosure requirement? Our AI Policies directory covers 150+ universities with verified L/D/E codes. Look up any school to see whether they require, encourage, or have no mechanism for AI disclosure.

---

Related Reading

• Do Top 10 Colleges Check for AI? Official Policies
• Do Colleges Use AI Detectors? The Truth About Turnitin
• How to Write an AI Disclosure for Your College Application
• Which Colleges Use AI in Admissions?
• College AI Detection Tools: The Spending Truth

---

Sources: Kaplan 2025 Admissions Officer Survey, Cornell University AI Essay Study (2025), foundry10 AI and College Applications Research, GradPilot AI Policies Database (150+ schools, accessed February 2026).