School AI policies, verified

“Brown affirms this definition and reiterates our own independent policy that the use of artificial intelligence by an applicant is not permitted under any circumstances in conjunction with application content. While an applicant may use artificial intelligence to assist with spelling and grammar review, in the same way as any other platform that supports basic proofreading, the content of all essays, short-answer questions and any other material submitted by an applicant must be the work of that individual.”

Brown's prohibition covers any AI use in conjunction with application content, with only a spelling-and-grammar exception. An AI panel reading and critiquing the essay is AI used in conjunction with application content, so CollegeEssayPanel reads this conservatively as prohibiting AI feedback.

Verified 2026-10-16 · published policy

“[Ethical examples:] 'Using AI tools, like Grammarly or Microsoft Editor, to review grammar and spelling of your completed essays' … 'Generating questions or exercises to help kick start the brainstorming process' [Unethical examples:] 'Copying and pasting directly from an AI generator' … 'Relying on AI generated content to outline or draft an essay' … 'Replacing your unique voice and tone with AI generated content' … 'Translating an essay written in another language'”

Caltech presents its guidance as Q&A headings with example bullets; the bracketed labels above are ours, the bulleted phrases verbatim. The ethical examples cover grammar and spelling review of completed essays, brainstorming kick-starts, and process research. Substantive critique of a completed draft is not addressed either way, so this reads narrower than a general feedback permission.

Verified 2026-10-16 · published policy

“The personal insight responses should reflect the student's own ideas and experiences, be independently written by the student and reflect the student's personal writing style. Students may receive advice on content and editing, including the use of generative artificial intelligence software to assist with readability, but content and final written text must be their own.”

One system-wide Statement of Application Integrity covers every UC campus, including UCLA and Berkeley. It permits advice on content and editing, names generative AI only as help with readability, and requires the content and final written text to be the student's own. It does not say whether AI may advise on content, and some guides read it as limiting AI to readability. A review that writes none of your text is advice under the statement's terms; if you read UC's line more narrowly, keep AI to readability for your UC answers.

Verified 2026-10-16 · published policy

“Using an AI platform to review one's grammar or spelling, or to seek general advice or topic suggestions at the start of the writing process does not constitute application fraud. Submitting personal statements or other written application responses composed by text-generating software may result in admission revocation or expulsion.”

Yale maintains a dedicated admissions AI-policy page with an explicit carve-out: grammar review, general advice, and topic suggestions are fine; submitting AI-composed text is application fraud. The two quoted sentences come from separate paragraphs of that page.

Verified 2026-10-16 · published policy

“A student's essay should reflect their own content, writing style, and English proficiency. … The Common Application identifies the following as a possible form of fraud and violation of its application standards: '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.'”

Harvard publishes the Common Application fraud standard on its own admissions FAQ without an explicit feedback carve-out. The prohibition targets submitting AI's substantive output as your own work (AI-written text) rather than AI review, so feedback reads as permitted; this is the strictest of the fraud-standard schools. Some outside lists count Harvard among schools that ban AI help; Harvard's own page adds no AI rule beyond the Common Application standard it quotes.

Verified 2026-10-16 · published policy

“It is Hopkins' policy that you may use AI tools for brainstorming or refining ideas before the writing process begins and for basic proofreading of your own writing, such as spelling and grammar review, but all material submitted within the application must be your own original work.”

Johns Hopkins' admissions FAQ explicitly permits brainstorming, idea refinement, and basic proofreading, and treats submitting AI's substantive content as application fraud under Common Application policies.

Verified 2026-10-16 · published policy

“Just as you wouldn't ask another person to write your essays for you, we expect you won't ask AI to do that work on your behalf.”

Northwestern maintains a dedicated admissions AI-advice page (updated 9/23/25). Its Constructive Uses of Generative AI list includes, as bullets, brainstorming essay topics and reviewing the grammar, punctuation, and spelling of your completed essays; problematic uses are AI-written, AI-translated, or AI-reworked essays.

Verified 2026-10-16 · published policy

“If you choose to use generative AI in your application process, we expect you to do so ethically. Ethical uses of generative AI include researching colleges, brainstorming essay topics, and reviewing the grammar and spelling of your completed essays. Unethical uses include (but are not limited to) using generative AI to outline, draft, or write your essays…”

Cornell's enrollment FAQ draws an explicit line: brainstorming and review of completed essays are ethical; outlining, drafting, or writing with AI is not. The quoted unethical-uses list continues with translation and portfolio images, unrelated to essay feedback.

Verified 2026-10-16 · published policy

“Students may utilize a wide range of resources, tools and support during the college admissions process. Regardless of the tools and support you may have used, we expect that all submitted application materials are your own original work and reflect your authentic voice.”

Columbia does not enumerate permitted or banned AI uses; it permits a wide range of tools while holding submissions to an own-original-work, authentic-voice standard, which excludes AI-written text but is consistent with AI feedback.

Verified 2026-10-16 · published policy

“We understand that you might use other forms of assistance, such as ChatGPT or other forms of AI, as you prepare your essay. … Likewise, AI should never be used to replace your own independent thinking. As you complete the essay portions of the application, you should always use your own voice and write about your own life experiences.”

Vanderbilt names ChatGPT directly and accepts AI assistance for brainstorming and grammar by analogy to human helpers (the elided text endorses asking a parent or peer for feedback on clarity and structure), while prohibiting AI that replaces the student's own thinking or rewrites the essay.

Verified 2026-10-16 · published policy

“AI should never replace your unique voice, experiences and personal expression and, if used at all, should only serve as a supplementary tool to enhance your writing (i.e., grammar and spelling checks, suggested structural or vocabulary improvements).”

Carnegie Mellon's admissions FAQ permits AI as a supplementary tool (grammar, spelling, structural and vocabulary suggestions) while urging caution and barring AI from replacing the applicant's voice.

Verified 2026-10-16 · published policy

“Your essays should tell your story, in your words. AI tools can be helpful when used thoughtfully, such as for brainstorming, outlining, proofreading, or organizing information. We recognize these tools may be especially useful for applicants who have limited access to application guidance. However, your application should reflect your own ideas, experiences, and authentic voice. Do not copy and paste AI-generated text or submit work that does not represent you.”

New for the 2026-27 cycle: Michigan replaced its fraud-standard-by-reference page with explicit guidance (Representing Your Authentic Voice, on the essay-questions page). The helpful-uses list is expressly non-exhaustive (such as), bounded by the authentic-voice requirement and a ban on pasting AI-generated text. Verified in a real browser 2026-08-04; the admissions domain blocks automated fetching.

Verified 2026-10-16 · published policy

“Reviewing the grammar and spelling of your completed essays can be a helpful use of AI, but you may ask a teacher or counselor to review as well. Unethical uses include (but are not limited to) using AI to brainstorm essay topics or draft or write your essays.”

Narrower than most permits-feedback schools: WashU allows grammar-and-spelling review of completed essays but explicitly lists brainstorming essay topics (not just drafting) among prohibited uses.

Verified 2026-10-16 · published policy

“While AI and similar tools can provide some assistance and feedback, it is essential that your thoughts, voice, and story remain at the core of your application. AI tools could polish or refine, but they should not replace or generate your personal input or creativity.”

Emory's admissions FAQ explicitly allows AI assistance, feedback, polishing, and refining, while prohibiting AI that replaces or generates the applicant's own content.

Verified 2026-10-16 · published policy

“Generative AI may play a minor role in your final product by helping you brainstorm essay topics or check for grammar mistakes, but we ask that you do not cut and paste any part of your essay or personal statements from an AI tool. All final submitted materials should be your original work.”

UVA permits brainstorming and grammar checks, prohibits pasting AI text, and backs the rule with the honor statement every applicant signs pledging the materials are not primarily a product of AI. Verified in a real browser 2026-08-04; the admission subdomain blocks automated fetching.

Verified 2026-10-16 · published policy

“While applicants may utilize AI as a tool during the college application process, their final application submission should be their authentic creation. NYU does not use AI detection technology as part of the review process.”

The most permissive verified policy: NYU affirmatively allows AI as a tool and asks only that the final submission be the applicant's authentic creation.

Verified 2026-10-16 · published policy

“Tufts Admissions strongly discourages the use of generative artificial intelligence (AI) for writing college admissions essays and short answer responses. … The Admissions Committee expects to read an applicant's original ideas presented in their own words.”

Tufts strongly discourages AI for writing essays; the hard prohibition is on submitting AI-generated text as your own. Discouragement is not a ban on feedback.

Verified 2026-10-16 · published policy

“We believe there is a place for them in helping you generate ideas, but your ultimate submission should be your own. As with all other sources, you should not copy and paste content you did not create directly into your application. … Use it to brainstorm, edit, and refine your ideas.”

The most affirmatively AI-friendly verified stance: Georgia Tech's admissions guidance encourages collaborative use of AI to brainstorm, edit, and refine, prohibiting only copy-pasted AI text.

Verified 2026-10-16 · published policy

“Getting help is not the same as losing ownership. Talking through an idea with a parent, teacher, or counselor, having someone proofread your writing, or using a tool, including an AI tool, to organize your thoughts or check your grammar, doesn't compromise your application certification. What would compromise your application is submitting work, ideas, or experiences that aren't authentically yours, whether that means someone else wrote your essay, an AI tool generated your responses, or your activities and honors are embellished or invented.”

New since the July sweep: Penn published a dedicated integrity-and-AI page that analogizes AI help to a parent, teacher, or counselor. Its enumerated AI examples are organizing thoughts and grammar checks, with the ban aimed at AI-generated responses; its own test is whether you could explain your writing in an interview. Verified in a real browser 2026-08-04.

Verified 2026-10-16 · published policy

“While an applicant may use artificial intelligence to assist with spelling and grammar review, or brainstorming support before writing, the content of all essays, short answer questions, activities, and any other material submitted must be an applicant's own work. Submitting content generated, drafted, or substantially written by artificial intelligence tools or other text-generating software may result in the revocation of your offer of admission or potential expulsion.”

New since the July sweep: Notre Dame published a dedicated AI Policy page under Apply that adopts the Common App fraud standard and enumerates its own permissions, which are spelling and grammar review plus brainstorming support before writing. Substantive critique of a completed draft is not addressed either way, so this reads narrower than a general feedback permission. Verified in a real browser 2026-08-04.

Verified 2026-10-16 · published policy

“Whether you are getting help from AI or another person, we recommend that any assistance with writing comes at the beginning or end of the process — such as brainstorming ideas for essay topics or asking for feedback on readability, spelling, and mechanics. [Not acceptable:] 'Asking AI or another person to outline or draft an application essay' … 'Using AI to modify the tone of your writing (e.g., rewriting to sound more academic or conversational)' … 'Asking AI or another person to rewrite a significant portion of your essay(s)' … As you would when quoting or citing work from human creators, the submission of any AI-generated writing, audio, video, or graphic content in your application must include proper acknowledgement or citation.”

Swarthmore publishes dedicated admissions guidance on AI use. The recommended examples cover brainstorming and feedback on readability, spelling, and mechanics; the list of work that counts as substantially created by AI or another person covers outlining, drafting, translating, rewriting significant portions, and changing the tone. Substantive critique of a completed draft sits between those lists, so this reads narrower than a general feedback permission, and any AI-generated content submitted requires citation. The bracketed label is ours; the page introduces the list as examples of work you are not submitting.

Verified 2026-10-16 · published policy

“[Permitted:] 'brainstorm ideas or reflect on possible essay topics' … 'review grammar or provide feedback after you have written a draft' [Prohibited:] 'you should not use AI to draft, write, or generate your responses' … 'you should not use AI to rewrite portions of your work or modify your tone' … 'you should not use AI or another tool to translate your written work from a different language into English'”

Bowdoin's dedicated AI-policy page is the clearest feedback permission in this directory: applicants may use AI to review grammar or provide feedback after writing a draft, while AI drafting, rewriting, tone modification, and translation are prohibited. The bracketed labels are ours; the quoted phrases are verbatim. Bowdoin applies the same ethic to human helpers.

Verified 2026-10-16 · published policy

“Generative artificial intelligence tools like ChatGPT may also serve as a resource, but must only be used as a guide; content must never be directly copied from AI or other sources. The submission of original content not only adheres to the ethical expectations in applying to college, but benefits the applicant by having their authentic personal story valued by the Admission Committee.”

Boston College's note on AI sits inside its application-requirements page: AI may serve as a resource and guide, which covers feedback on an applicant's own writing. The line is drawn at copying AI-produced content into the application. Every submitted word must be the applicant's own.

Verified 2026-10-16 · published policy

“If you opt to use these tools at any point while writing your essays, they should only be used to support your original ideas rather than to write your essays in their entirety. As potential future Terriers, we expect all applicants to adhere to the same standards of academic honesty and integrity as our current students. When representing the words or ideas of another in their original work, students should properly credit the source.”

Boston University's undergraduate Statement on Generative AI permits AI tools that support the applicant's original ideas and prohibits having AI write the essay. Feedback on a draft is a supporting use. BU also notes the most compelling personal statements carry the writer's authentic voice. Some BU graduate programs publish stricter rules; this entry covers undergraduate admissions.

Verified 2026-10-16 · published policy

“The Admission Committee expects that each writing supplement be the original work of the applicant, unaided by artificial intelligence. Those found to be submitting writing that is not their own may be denied or have their admission rescinded.”

Villanova's Important Note names the writing supplements specifically: each must be 'unaided by artificial intelligence,' with denial or rescission as stated consequences. 'Unaided' admits no feedback exception, so CollegeEssayPanel reads this as prohibiting AI review of Villanova-bound essays. The note sits on the page that also carries the Common App and Coalition essay requirements without a separate AI statement for them; we apply the conservative reading to any essay named for Villanova rather than guessing where the line sits. One 2026-27 supplement prompt opens by saying Villanova embraces AI with thoughtful, ethical use and asks how technology can help you serve the common good; that is a question about technology, not permission to use AI on the supplement, and the 'unaided' requirement sits on the same page.

Verified 2026-10-16 · published policy

“The rise in artificial intelligence tools (e.g. ChatGPT) creates the ability to use shortcuts that may seem attractive, but won't yield the same result as reflection, revision, and original thought. … AI can't capture how you think or what you value. You can use AI as a tool—but proceed with caution. … Using AI in its proper place, such as brainstorming ideas or polishing a draft, ensures that your work remains original.”

Fordham publishes a dedicated AI Policy for Application Writing page that names brainstorming and polishing a draft as AI 'in its proper place' while warning that AI-manufactured content defeats the essay's purpose. Feedback on the applicant's own writing sits squarely inside that permission. The page is public but served through Fordham's login gateway, which passes anonymous visitors through; verified in a real browser 2026-08-09.

Verified 2026-10-16 · published policy

“While artificial intelligence tools may be helpful in generating essay topic ideas, and in editing or refining students' writing, the University of Rochester discourages the use of AI to compose application essays or short answer responses. The effect often stifles an applicant's authentic voice – a key element of what we hope to hear within their personal statement.”

Rochester's statement, published with its application instructions, itself acknowledges AI can help with topic ideas and with editing or refining a student's writing, and stops at discouraging (not prohibiting) AI-composed essays. Feedback on an essay the student wrote is on the permitted side of that line.

Verified 2026-10-16 · published policy

“We will not disqualify an applicant found to have used or suspected of using AI in their admissions essays, nor are we running essays through any system to detect if AI was used. However, we strongly discourage students from simply feeding AI a prompt for their essay, as it denies you the opportunity to help us better understand why you would be a good fit for UW–Madison.”

UW–Madison's essay page carries a dedicated 'Use of AI in Admissions Essays' section and is the most explicit non-enforcement statement in this directory: no disqualification for AI use and no AI detection, paired with a strong discouragement of wholesale AI generation. Feedback on a student-written essay is not discouraged by the text at all.

Verified 2026-10-16 · published policy

“All writing in the application, including your essay/personal statement, must be your own original work. If you choose to have a parent, counselor, tutor, friend or AI tool review your writing, it must be done responsibly and ethically — again, it must be your own original work.”

The University of Washington's first-year writing-section page names an 'AI tool' alongside parents, counselors, tutors, and friends as an acceptable reviewer of the applicant's writing, conditioned on responsible use and the essay remaining the applicant's own original work. It is the most direct permission for AI feedback in this directory.

Verified 2026-10-16 · published policy

“May I use artificial intelligence (AI) to help write my admissions essay? AI tools may be used like any other general writing aid, but the essay submitted must be your own work. The goal is for UGA to learn about you in your own voice.”

UGA answers the question directly in its first-year FAQ: AI tools may be used like any other general writing aid, with the requirement that the submitted essay be the applicant's own work, in their own voice. A general writing aid covers feedback and editing help.

Verified 2026-10-16 · published policy

“When you apply, you'll agree to an Essays and Activity Commitment, which includes: … Not using paid help or using generative AI tools (like ChatGPT). … Using your own words and ideas (you may ask others to review your essays for clarity).”

BYU's application includes an Essays and Activity Commitment that every applicant agrees to, and its first item is not using generative AI tools such as ChatGPT. The page lets other people review essays for clarity and extends no such allowance to AI, so CollegeEssayPanel reads the commitment as prohibiting AI review, the same conservative reading applied to Brown. The commitment replaced an earlier statement, quoted here until October 2026, that barred generative AI while composing and warned that offers could be rescinded for AI-generated essays; the new wording is broader, not narrower. The commitment's items are a bulleted list, joined above with ellipses. Applies to BYU's general application essays, not an honors track.

Verified 2026-10-16 · published policy

“However, your ultimate submission should be your own. As with all other sources, do not copy and paste content you did not create directly into your application. If you choose to utilize AI-based assistance while working on your writing submissions for UConn, we encourage you to take the same approach when collaborating with people – use it to brainstorm, edit, and refine your ideas.”

One of the most explicit permissions in the directory: UConn names brainstorming, editing, and refining as acceptable AI uses and asks only that the submitted words be the applicant's own. Published on the main Apply page and applying to first-year admission generally.

Verified 2026-10-16 · published policy

“If you choose to use AI, it should be limited to basic proofreading, such as correcting grammar, punctuation, and spelling errors.”

CMC's first-year application instructions carry a section headed Essay Support and Artificial Intelligence, so this is deliberate admissions guidance rather than a general integrity rule read across. It admits AI for spelling, punctuation, and grammar and stops there, and separately says the essays should be written by the applicant. A panel review is substantive feedback, which sits outside that limit, so the conservative reading applies here as it does at Brown.

Verified 2026-10-16 · published policy

“Further, the applicant understands that the use of artificial intelligence (AI) tools to complete any portion of the application, including essays, is prohibited.”

Georgetown's first-year application ends with a signed attestation, readable on the public application form, in which every applicant affirms that using AI tools to complete any portion of the application, including essays, is prohibited. It makes no exception for grammar or feedback. Because it is worded around completing the application, a narrower reading (AI that writes) is arguable; CollegeEssayPanel does not take it, for the reason it does not at BYU: the student signs this statement, and being wrong costs them the application. Until October 2026 this attestation was known only from secondary reports and listed here as an unverified caution.

Verified 2026-10-16 · published policy

“As part of our contextual review at Rice, we value your authentic voice and story above all else. AI can be a helpful tool for brainstorming or organizing your ideas, much like consulting teachers, counselors, or family members. However, preserving your original thoughts remains essential. In alignment with Rice’s commitment to responsible AI usage, any resource you use should ultimately amplify your voice and should not create content for you, as no one else can articulate your unique lived experiences and perspective.”

New since the August sweep, when Rice published nothing: its undergraduate admission FAQ now likens AI to consulting teachers, counselors, or family members for brainstorming and organizing ideas, and draws the line at any resource creating content for you. The FAQ does not name feedback either way; feedback on an essay you wrote amplifies your voice without creating content, so it sits on the permitted side of that line.

Verified 2026-10-16 · published policy

“A candidate’s personal statement must be their own work. The personal statement’s purpose is to provide more insight into the candidate’s own ideas, background, and writing abilities in English. Wesleyan recognizes there are artificial intelligence (AI) tools and resources available to candidates. Nonetheless, we expect that candidates’ responses to all application questions are authentically their own and reflective of their abilities. Candidates may not submit as their own, any content either copied and pasted from an AI resource or authored by another individual. If it is determined that an applicant received inappropriate assistance with the application essays, candidacy will be compromised and/or admission rescinded.”

Wesleyan publishes this under a Special Statement About Authenticity, AI, and the Use of Agents on its International Applicants page. The wording addresses candidates generally and the other admission pages we checked carry no separate statement, so it is listed here for every Wesleyan-bound essay. It bars submitting content copied from an AI resource as your own and warns that inappropriate assistance with the essays can cost the application or the offer. It does not say whether AI feedback counts as inappropriate assistance; a review that writes none of your text leaves the submitted words yours, which is the line the statement draws. Some outside lists count Wesleyan among schools that ban AI help; its own statement bans AI-written content and does not mention feedback.

Verified 2026-10-16 · published policy

“Plenty of applicants will use them somewhere in their writing process from brainstorming a topic, to getting unstuck, to improving a sentence that isn’t landing. Those examples aren’t too different from asking a teacher or a parent to read a draft, and we don’t expect you to pretend these AI tools don’t exist. What we do expect is that the words and thoughts you share are yours. If AI is drafting your essay, rewriting your paragraphs, or generating the ideas you’re presenting as your own thinking, that crosses a line, even if you technically “wrote” the final version by copying it down.”

New since the August sweep, when USC published nothing: its undergraduate What We Look For page added a note on AI covering the Common App essay, USC's short responses, and the activities list. It likens brainstorming, getting unstuck, and improving a sentence to asking a teacher or parent to read a draft, and draws the line at AI drafting the essay, rewriting paragraphs, or generating the ideas you present as your own. USC also says it does not penalize thoughtful, appropriate AI use. Feedback on a draft you wrote sits on the accepted side of that line.

Verified 2026-10-16 · published policy

“Good writing is a process, and there are many resources you might use as you craft your responses, including asking someone you trust to review your work and offer feedback or using generative artificial intelligence to brainstorm your response. But please know that what is most important to us is to hear your voice and ideas.”

Haverford's first-year supplement page, which every Early and Regular Decision applicant completes, names generative AI as a resource for brainstorming responses and credits feedback to someone you trust, without saying whether AI feedback counts. Nothing on the page restricts AI, and it asks above all to hear your voice. The statement covers the Haverford supplement responses and does not separately address the Common App essay. It has been on the page since at least January 2024; this directory listed Haverford as publishing nothing until October 2026. Verified in a real browser, because the site blocks automated retrieval.

Verified 2026-10-16 · published policy