Purdue academic integrity and AI: a 1972 definition, and what to ask for

The short answer

Purdue’s published integrity regime does not name artificial intelligence anywhere, and the language it does use was written in 1972 — so the rule you are actually graded against is the one in your syllabus. Get it in writing, and keep the reply.

This is a short page, deliberately. Several of the questions students most want answered — is AI detection switched on, is there a transcript notation, what standard of proof applies — have no published answer at Purdue. We say so rather than filling the gap with plausible-sounding detail, and the last section is about what to ask for instead.

The definition, and its date

Purdue does not have a single named “Academic Integrity Policy” regulation. The operative text sits in Student Regulations, Part 5, Section III-B-2-a, quoted by the Office of Student Rights and Responsibilities:

“dishonesty in connection with any University activity. Cheating, plagiarism, or knowingly furnishing false information to the University are examples of dishonesty”

Alongside it, the same page quotes the University Senate — and this is the sentence worth holding on to, because of the date attached to it:

“the commitment of acts of cheating, lying, and deceit in any of their diverse forms…is dishonest and must not be tolerated”
— University Senate Document 72-18, December 15, 1972

Fifty-plus years is not a criticism in itself; a broad principle can outlive its technology. But it does tell you where the specifics live, and it is not in the policy. No revision date is shown on the OSRR page at all.

Plagiarism is defined more concretely. It occurs when

“one person steals another person’s ideas or words and falsely presents them as the plagiarist’s own product”

— including using exact language without quotation marks or attribution, presenting others’ ideas or arrangements without acknowledgment, and submitting documents written by others as one’s own.

That third limb is the one an AI allegation would most naturally run on: submitting documents written by others as one’s own. Whether a model counts as an “other” is exactly the question the 1972 language was never asked.

The Honor Pledge, separately published, reads:

“As a Boilermaker pursuing academic excellence, I pledge to be honest and true in all that I do. Accountable together—We are Purdue.”

That page states no approving body and no adoption date; it carries “Last modified May 14, 2025”.

AI is not named anywhere

No mention of AI or generative AI appears in Purdue’s academic-integrity text. The definition is technology-neutral throughout.

Purdue does run an AI teaching hub through Innovative Learning, but what it publishes is workshops and courses — an “AI Bytes” series, and self-paced AI courses new for fall 2026. It contains no statement on AI detection, on Turnitin AI detection, on AI syllabus statements, or on academic integrity. No published position on any of those was located.

Nor is there a university-wide AI-disclosure or citation requirement: none is publicly documented. If your instructor wants a disclosure, their format is the format. Our AI disclosure statement guide gives you wording to adapt when nobody hands you a template.

Turnitin Originality, and what is not said about it

Purdue names its tool plainly:

“Turnitin Originality is Purdue’s centrally supported originality checker and plagiarism detection tool”

That page is published by Innovative Learning, supported by Purdue Information Technology (it@purdue.edu, 765-494-4000), and carries “Last modified: February 18, 2026”. The tool is enabled per assignment in Brightspace, so whether your work is run through it is an instructor decision, course by course.

(Purdue’s tool index lists it as “Turnitin Originality — Plagarism checker”. The misspelling is in the original.)

Whether the Turnitin AI-writing indicator is switched on for Purdue is not publicly documented. A page last modified in February 2026 that says nothing about AI writing detection either way is genuinely silent — not evidence that it is off, and not evidence that it is on.

Worth separating the two things that page does and does not cover: an originality checker answers “does this text match text we have seen before”. An AI indicator answers “does this text look machine-written”. They are different measurements with different failure modes, which is the subject of what a Turnitin score actually means.

There is no published student self-check

Not publicly documented. Draft Coach is not mentioned on Purdue’s Turnitin page, and while that page links Turnitin’s own student guidance on reading a similarity score, it describes no student-initiated checking route.

Repository storage is in the same position: whether Purdue submissions enter Turnitin’s comparison repository, for how long, and whether any opt-out exists, is not addressed on that page. If your thesis is unpublished and that matters to you, the instructor who enabled the assignment is the person who can answer.

Sanctions, and the questions with no published answer

The published sanction list is short:

“a warning, probation, probated suspension, suspension or expulsion”

Separately, instructors may impose punitive grading — lower grades, course failure, or requiring reassignment completion. In practice that is the sanction most students meet, because the instructor can apply it directly.

On process, what Purdue publishes is this much:

• The instructor investigates first, and may resolve the matter by punitive grading.
• Cases may be referred to the Office of the Dean of Students for formal adjudication under Part 5, Section III of Student Regulations.
• On appeal, a “grade appeals system offers recourse to a student whose grade has been reduced unfairly for alleged academic dishonesty.”

And here is what is not published, stated plainly:

• No transcript notation. No XF-style notation is published on the OSRR academic-integrity page. That is an absence of documentation, not a promise that no record exists.
• No standard of proof.The page does not say whether a preponderance standard, or any other, applies. Most US institutions publish one; Purdue’s integrity page does not.
• No published case data. No counts, no outcomes, no rates. Nobody can tell you how common any of this is at Purdue.

None of that means you are without recourse. It means the recourse is not written down in advance, so you have to ask for it.

What to ask for, in writing

A thin rulebook cuts both ways. It gives an instructor wide discretion — and it gives you room to ask precise questions that the published material does not foreclose. If you are accused, these are the four:

• What is the allegation, specifically? Which assignment, which passages, and what makes them the concern.
• What is the evidence? If it is a tool output, which tool, and what the number is being taken to mean. A similarity score and an AI score are not interchangeable.
• What standard is being applied? Purdue does not publish one for integrity cases, so it is fair to ask what threshold the decision-maker is using.
• What is the route and the deadline? Whether this is being resolved by punitive grading or referred to the Dean of Students, and how the grade appeals system applies to your case.

Then bring your process: drafts, dated files, notes, search history, outlines. How to prove you wrote it yourself is the practical version of that, and what to do when you are accused covers the first 48 hours. If you write in English as a second language, AI detectors and non-native speakers is the research to have with you.

If a detector score is what started this, our guide to an AI detector flagging your thesis sets out what that number does and does not establish on its own.

Before you submit

• Ask your instructor what AI use is permitted, and keep the answer. Purdue’s policy is silent, so the syllabus and that email are the rule.
• Assume Turnitin Originality may be on for the assignment. It is centrally supported and enabled per assignment in Brightspace.
• Quote and attribute exactly. The plagiarism definition reaches exact language without quotation marks, and ideas or arrangements presented without acknowledgment.
• Keep dated drafts. Where a regime publishes little, your own evidence carries more weight, not less.
• Check that every citation exists and says what you claim. See how to verify your citations.

If you want a read on your own text before you hand it in, our AI check is free up to 1,500 words, with no account and no name. Our published error rates are measured on English and German corpora, and no detector — ours included — produces proof.

Sources

• Purdue Office of Student Rights and Responsibilities: Academic Integrity — quotes Student Regulations Part 5, Section III-B-2-a and University Senate Document 72-18 (15 December 1972); the plagiarism definition, the sanction list and the grade appeals sentence. No revision date shown.
• Purdue OSRR: Honor Pledge — carries “Last modified May 14, 2025”; states no approving body or adoption date.
• Purdue Student Regulations: University Regulations — Part 5, Section III is the formal adjudication route.
• Purdue Innovative Learning: Turnitin — “Purdue’s centrally supported originality checker”; page “Last modified: February 18, 2026”; supported by Purdue IT; enabled per assignment in Brightspace.
• Purdue Innovative Learning: AI tools and resources — workshops and courses only; no statement on AI detection or academic integrity.

This page is orientation, not legal advice. What binds you is Purdue Student Regulations Part 5 and your course syllabus.