Free review

School AI policies · Methodology

How the directory is built

As of October 16, 2026: 5 of the 153 schools we checked publish a policy that rules out AI feedback on application essays; 36 publish a policy that stops short of a ban, most of them permitting feedback and prohibiting AI-written text; 112 publish no policy we could verify. This page says how those numbers are produced: which schools are in the directory, what each tier means, what counts as a source, how often every entry is re-read, and how to cite or reuse the data.

Which schools

The directory began with the thirty schools applicants name most and now covers every state: the public flagships, the selective private universities and liberal arts colleges, and three state systems (the University of California, the California State University and SUNY) as single entries, because the system rather than the campus publishes the rule. It holds 153 schools as of October 16, 2026. Schools are added as applicants to the panel name them; none has been removed.

The three tiers

Verified. The school publishes a rule on applicants' use of AI on a page it controls. We quote the operative sentence verbatim, link the page, and record the date we read it. The stance is read from the quote: either the school prohibits all AI involvement, feedback included, or it stops short of that and prohibits AI-written text while leaving feedback alone or within limits. Where a sentence can be read two ways, we record the stricter reading, and the label on the page says exactly what the quote supports and no more. 5 schools prohibit all AI involvement; 36 stop short of a ban.

Reported, not publicly verifiable. A policy described by a secondary source, such as a news article, a counselor's post, or an attestation seen only behind a login, that we could not find on the school's own public pages. It is shown as a caution with the date we looked, it never stops a review, and it moves to the verified tier the day the text becomes readable without an account, as Georgetown's did.

No published policy. We read the school's admissions pages and found no rule on applicants' use of AI. The page records the date. For the 48 schools searchers ask about most, it also lists every page we read, the applications the school accepts, and what the school does say about the essay and about honesty, in its own words. Silence is not permission: the Common App rule below still applies, and so does the school's general integrity policy.

What counts as a source

A page on the school's own domain, or on a domain the school owns (MIT's admissions site is mitadmissions.org), readable without logging in. Text inside the application form counts once it can be read without an account. News coverage, counselor blogs, forums and aggregator sites can prompt a check, but they are never quoted and never move a school between tiers. We do not paraphrase a quote, cut it with an ellipsis that changes its meaning, or read an academic-integrity policy written for enrolled students as a rule for applicants.

The rule every Common App applicant signs

Most schools in the directory accept the Common Application, whose fraud policy (last updated August 2023) lists among its definitions 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;”

and every applicant signs an affirmation on submission that begins:

“I certify that all information submitted in the admission process — including this application and any other supporting materials — is my own work, factually true, and honestly presented”

Read November 1, 2026 · Common App fraud policy · Common App application affirmations

Neither sentence mentions feedback. That is the line most verified policies draw too: help that leaves you the author is not the same act as submitting text you did not write. Where a school says nothing, that distinction is yours to make and yours to verify.

How often we check

Every entry is re-read against its source at least once per admissions cycle, and whenever a reader reports a change. The headline date on the directory is the oldest check date of any entry, so it never claims a sweep newer than the entry nobody re-read; each school's page carries its own date. Supplemental essay prompts are swept separately, last on October 24, 2026. Every change to the data is logged with the schools it touched on the changes page.

Corrections

If a school's policy has changed, or we have misread it, tell us with the page's address. We re-read the source, update the entry and its date, and log the change. We do not change a stance on a reader's word alone; we need the school's page.

Licence and citation

The directory is published under CC BY 4.0: reuse it, with attribution. Cite it as:

CollegeEssayPanel, "College AI essay policies, verified," https://collegeessaypanel.com/policies, as of October 16, 2026. CC BY 4.0.

Download: CSV · JSON

The per-school pages, the directory table and both downloads are generated from the same data, so a quote or a date never differs between them.

What we do not do

What the panel does with it

Target schools a student names are matched against the directory before a review starts. If a matched school's verified policy prohibits all AI involvement, including feedback, the review stops and the credit is returned. A reported policy is shown as a caution beside the review; a school with no published policy proceeds, with a reminder to verify. The directory is why the panel works the way it does: feedback that keeps the student the author, and nothing that writes for them.