Academic integrity and AI at the University of Minnesota: no grade notation, but a seven-year record

The short answer

Minnesota does not put a letter on your transcript. It keeps a disciplinary record and releases it to third parties for seven years — unless you petition to have it expunged before then. That petition is the action item on this page, and it is easy to miss because nothing on your transcript reminds you it exists.

The rest is more ordinary. The Board of Regents Student Conduct Code, last amended 10 June 2022, does not mention AI at all; it reaches the question through “unauthorized use” and plagiarism, both of which are decided by what your instructor permitted. And UMN has published no position on whether Turnitin’s AI detection is used — the substantive guidance that exists sits behind a university log-in.

Seven years, and the expungement petition

Start here, because it is the part that outlives everything else. The Office for Community Standards publishes this, verbatim:

“The University of Minnesota will release disciplinary records to third parties, if given permission by the student. For the Twin Cities campus, this Expungement Petition should be used to request an exception to this policy and have the disciplinary record changed to non-disciplinary before the seven years.”

“If an expungement is petitioned and granted, the student conduct record will be changed to non-disciplinary. As a result, the record of the violation will not be disclosed to external third parties by the Office for Community Standards (OCS) or Housing and Residence Life (HRL) unless you have granted permission or unless OCS is required by law.”

What that means in practice:

• A finding leaves a disciplinary record, not a grade mark. Nothing is branded onto the transcript itself.
• That record is released to third parties — the graduate programme, the professional licensing body, the employer running a background check — where you give permission, and many applications require you to.
• Seven years is the default duration. Expungement is the published exception, and it is something you have to apply for. It does not happen because time passed or because you graduated.
• Granted expungement changes the record to non-disciplinary, after which OCS and Housing will not disclose it to external third parties without your permission or a legal requirement.

Scholastic-dishonesty cases are routed through a specific process: “Violations of Student Conduct Code subdivision 1: Scholastic Dishonesty are addressed through the Academic Integrity Matters (AIM) process.”

One more thing on transcripts. There is no notation — but one of the published outcomes, “restriction of privileges”, is defined as “the denial or restriction of specified privileges, including, but not limited to, access to an official transcript for a defined period of time.” So the transcript can be withheld from you even though nothing is written on it.

What the Student Conduct Code actually says

The governing document is the Board of Regents Policy: Student Conduct Code, ten pages, category Academic, header stating “Last Amended: June 10, 2022”. The approving body is the Board of Regents of the University of Minnesota — the highest one available, which is worth knowing when someone tells you a rule is “just departmental”.

Its scope: “This policy applies to all students and student groups at the University of Minnesota (University), whether or not the University is in session.”

Here is the whole operative definition, from Section VI, Disciplinary Offenses, General Behavior Prohibitions, Subdivision 1, Scholastic Dishonesty:

“Scholastic dishonesty means plagiarism; cheating on assignments or examinations, including the unauthorized use of online learning support and testing platforms; engaging in unauthorized collaboration on academic work, including the posting of student-generated coursework on online learning support and testing platforms not approved for the specific course in question; taking, acquiring, or using course materials without faculty permission, including the posting of faculty-provided course materials on online learning support and testing platforms; submitting false or incomplete records of academic achievement; acting alone or in cooperation with another to falsify records or to obtain dishonestly grades, honors, awards, or professional endorsement; altering, forging, misrepresenting, or misusing a University academic record; or fabricating or falsifying data, research procedures, or data analysis.”

Notice how much of it is about posting: uploading your own coursework, or your instructor’s materials, to an outside platform is named twice, independently of whether you got anything back. If you have ever pasted a whole assignment brief into a service to get help, that clause is the one to read again.

AI is not named — and what reaches it instead

The strings “artificial”, “AI” and “ChatGPT” do not occur anywhere in the Student Conduct Code. The 2022 amendment language reaches at “online learning support and testing platforms” — that is the homework-answers-site problem, not the generative-AI one.

So an AI allegation at Minnesota has to run on one of two things: plagiarism, or unauthorized use. Both are settled by what your instructor permitted, which is why the syllabus does more work here than the Regents policy does.

The administrative policies say the same thing from the other side. The student-responsibilities policy (which excludes the Duluth campus, and carries no visible effective date in its body) states:

“Maintaining academic integrity.Students are expected to maintain academic integrity, including doing their own assigned work for courses. When it is determined that a student has engaged in scholastic dishonesty, the instructor may impose an academic consequence (e.g., giving the student a grade of ‘F’ or an ‘N’ for the course), and the student may face additional sanctions from the University.”

The instructor-responsibilities policy contains no mention of AI, generative AI, Turnitin or plagiarism-detection services anywhere in its text. What it does require of instructors is worth knowing, because it gives you something to ask for:

“Instructors should discuss scholastic dishonesty and what it means in the context of their class (e.g., define collaboration and when it is permitted, requirements about and methods for citing sources, receiving or giving aid on tests, and using electronic aids or communications during exams when prohibited).”

If your course has not defined what counts, asking for that definition is asking your instructor to do something the policy already says they should. (The effective date and approving body of that policy were not visible in the page body as we read it, so we do not state them.)

UMN’s position on AI detection: routed, and login-gated

This section is mostly about what is not published, and the distinction between two different kinds of missing.

Whether Turnitin’s AI-writing detection is used at Minnesota is not publicly documented. The IT@UMN Turnitin service page contains no occurrence of AI detection anywhere in it. The Office for Community Standards names Turnitin to instructors without reference to AI. No decision, either way, has been published.

The nearest published statement is explicitly a deferral rather than a decision. Minnesota’s Writing Across the Curriculum programme publishes a page on restricting ChatGPT use in classes with writing assignments — it carries no visible date — with a heading reading “AI Detection Platforms Generate False Positives and False Negatives”. Under it:

“OpenAI has recently discontinued its proprietary detection tool after determining that its false positive and false negative rates were unacceptably high. Free, proprietary models like ZeroGPT note that false positives and negatives remain common and that it is not always a reliable indicator of AI use. The Office of Information Technology can answer questions about the suitability of Turnitin as an AI detection tool.”

Those first two sentences are WAC’s own account of what those vendors said; we quote them as WAC’s, and we do not adopt any accuracy claim about any detector as our own. The sentence that matters institutionally is the last one: the question is routed to the Office of Information Technology, and OIT has published no answer. That is the whole of Minnesota’s public position.

A second WAC page, dated 5 January 2026, addresses false positives against multilingual writers specifically and cites Liang et al. (2023). It is a UMN teaching-blog post rather than policy — useful reading, not a rule. If you write in English as a second language, our own summary of that research is at AI detectors and non-native speakers, and the wider picture is in how accurate AI detectors really are.

Now the important distinction. Substantive UMN guidance on AI detection does exist — it is simply unreadable from outside, not absent. UMN’s teaching-support site names a Canvas course, “Artificial Intelligence and Academic Integrity: Resources for Instructors”, which “covers student engagement strategies, AI detection tools, citation practices, assignment design, and steps for responding to potential misuse”. Access requires a UMN log-in via TrainingHub, so its content cannot be quoted from a public source.

That has a practical consequence for you. If you are accused, your instructor may well be working from guidance you have never seen. Asking what guidance the department is applying, and asking to be shown it, is a fair and specific request — and a much better question than arguing about detectors in the abstract. If a case has already started, what to do when you are accused is the first thing to read.

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.

One access note, in the interest of not pretending to more than we have: every umn.edu host we tested returns HTTP 403 to ordinary automated requests behind a Cloudflare interstitial. The sources on this page were read in a normal browser session. Two URLs that circulate in older write-ups are dead — policy.umn.edu/education/academicdishonesty returns the Policy Library’s 404 page, and it.umn.edu/services/technology/turnitin returns IT@UMN’s 404.

Turnitin at Minnesota

Turnitin is in use, in two Canvas integrations. From IT@UMN, verbatim:

“Turnitin’s plagiarism prevention tool generates similarity reports with Canvas Assignments that show how much of a document is original, cited from other sources, or unoriginal.”

“The University features two ways to integrate Turnitin plagiarism-checking technology into your Canvas Assignments: Turnitin LTI 1.3 version… Canvas Plagiarism Framework (CPF) with Turnitin…”

The Office for Community Standards describes it to instructors in one line: “Turnitin is a grading and evaluation tool that assists in detecting plagiarism.”

Note the framing throughout: plagiarism, similarity. Similarity and authorship are different questions with different failure modes — see what a Turnitin score actually means.

There is no published student self-check route. Draft Coach is not mentioned on the IT@UMN Turnitin page, and both documented integrations are instructor-configured Canvas assignments. So you cannot see your own similarity report before submitting.

On the repository question: whether UMN submissions enter Turnitin’s comparison repository, for how long, and whether any opt-out exists, is not publicly documented. The IT@UMN page describes submission handling — “Only one file submission is allowed per assignment… All previous submissions will be retained and can be viewed through Canvas Speedgrader” — but says nothing about the repository. One adjacent line from the instructor-responsibilities policy is worth keeping: “Term papers and comparable projects are the property of students who prepare them (see Board of Regents Policy: Copyright.)”

Permission is per course, and does not travel

Minnesota has no university-wide AI disclosure or citation requirement. It is set per course, and IT@UMN’s FAQ for instructors and students (the page carries no visible date) is unusually clear about the boundary:

“Students who would like to use AI tools for coursework should always check with the instructor to understand expectations for each course.”

“How can a student find out if AI tools are allowed in their classes? Guidelines about appropriate use should be included in the syllabus for each class. Students should reach out to their instructors when they have questions about AI use for their course.”

“Does permission from one instructor to use GenAI extend to other classes or areas within the University? No, your instructor’s permission is specific to their own course and assignment expectations. It does not extend to other courses or override existing University policies.”

That last answer is the one students get wrong. Permission is not a status you acquire; it is granted assignment by assignment. A supervisor who is relaxed about using a model to tidy your English does not authorize anything in the seminar next door.

On citation, UMN does not specify a form: “If generative AI tools are allowed in coursework, how should they be cited? The University Libraries have helpful information and a tutorial on how to properly cite generative AI tools.” A pointer, not a template. If you need wording, our AI disclosure statement guide gives you something to adapt to whatever your instructor asks for.

And on reporting: “How should faculty report concerns about a student’s inappropriate use of AI tools? Information about how to respond to concerns about scholastic dishonesty, including inappropriate use of AI tools, can be found on the Office for Community Standards’ website.” The Office for Community Standards’ own student-facing page routes students to that FAQ in turn: “Review frequently asked questions about using generative artificial intelligence (AI) tools in your classes.”

The outcomes list

The Student Conduct Code PDF itself contains no sanctions list and no occurrence of the word “transcript”. The published scale is the Office for Community Standards’ “outcomes” list (the page carries no visible date):

• Academic Outcome — “an outcome affecting the course or academic work of the student for violation of Section VI, Disciplinary Offenses, Subdivision 1, Scholastic Dishonesty.” This is the one most scholastic-dishonesty cases end in.
• Warning — “the issuance of an oral or written warning or reprimand.”
• Disciplinary Probation — “special status with conditions imposed for a defined period of time and includes the probability of more severe disciplinary outcomes if the student is found to violate any institutional regulation during the probationary period.”
• Required Compliance — “satisfying University requirements, work assignments, community service, or other discretionary assignments.”
• Confiscation and Restitution — confiscation of goods or falsified identification; “making compensation for loss, injury, or damage.”
• Restriction of Privileges — “the denial or restriction of specified privileges, including, but not limited to, access to an official transcript for a defined period of time.”
• University Housing Suspension and Expulsion — temporary or permanent separation from University Housing.
• Suspension — “separation of the student from the University for a defined period of time, after which the student is eligible to return… Suspension may include conditions for readmission.”
• Expulsion — “the permanent separation of the student from the University.”
• Withholding of Diploma or Degree — “the withholding of diploma or degree otherwise earned for a defined period of time or until the completion of assigned outcomes.”
• Revocation of Admission or Degree.

For a thesis writer, “Withholding of Diploma or Degree” and “Revocation of Admission or Degree” are the two that reach past graduation. The degree is not final in the way it feels final.

Minnesota publishes no scholastic-dishonesty case counts or outcomes that we could locate, so nobody can tell you how often any of this happens.

Reporting, proof and one appeal

Two things happen when an instructor suspects scholastic dishonesty, and the second is not optional. The instructor may impose the academic consequence themselves — an F or an N in the course, per the student-responsibilities policy. And the instructor-responsibilities policy requires reporting:

“F. Report Scholastic Dishonesty — Instructors are obligated to report suspected scholastic dishonesty to their departments and to the appropriate office on campus (on the Twin Cities campus, the Office for Community Standards…).”

So a quiet arrangement with your instructor is not really available at Minnesota — the report is an obligation, and the AIM process follows from it.

The procedural policy, Resolving Alleged Student Conduct Code Violations, sets out what campus procedures must do. The list is worth reading as a list of your entitlements:

“allow the student to be accompanied or represented by an advisor of their choice throughout the student conduct process as permitted by campus procedures; provide timely notice of any meeting or proceeding; provide the range of possible outcomes; encourage informal resolution of alleged violations without the need for a hearing; permit students the opportunity for a formal hearing upon request; provide timely and equal access to information that will be used during disciplinary hearings; provide prompt and fair, unbiased resolution; provide for a preponderance of the evidence (i.e., more likely than not) standard of proof throughout the disciplinary process; and provide one campus-wide appeal of a finding of violation of the Code.”

Four of those are worth acting on:

• You may bring an advisor of your choice. Not everyone knows this, and going alone to a meeting is a choice, not a requirement.
• A formal hearing is available on request. Informal resolution is encouraged, but it is not the only route.
• You are entitled to see the information that will be used. If that information is a tool output, ask which tool and what its score is being taken to mean.
• The standard is preponderance of the evidence — more likely than not — and there is exactly one campus-wide appeal. One. So the appeal you file should be the considered one.

Each campus provides a hearing body, “comprising a hearing panel or hearing officer”, to conduct requested hearings.

Against a “more likely than not” standard, concrete process evidence is the most useful thing you can bring — dated drafts, outlines, notes, reading, the email where you asked what was allowed. That is the subject of how to prove you wrote it yourself.

Before you submit

• Ask per course, and keep the answer. Permission from one instructor explicitly does not extend to another course.
• Do not post coursework or course materials to outside platforms. The Conduct Code names posting twice, and it does not require that you received anything back.
• Keep dated drafts and notes. Minnesota gives you no way to self-check, so your own record is the evidence you will have.
• Verify every citation exists and says what you claim. Fabrication of “sources, citations, data, or results” is named in the Code — see how to verify your citations.
• If a finding is made, diary the expungement petition. Seven years of third-party release is the default; the petition is the exception, and nobody will file it for you.
• Remember you may bring an advisor to any meeting, and that a formal hearing is available on request.

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

• Board of Regents of the University of Minnesota: Board of Regents Policy: Student Conduct Code (PDF) — “Last Amended: June 10, 2022”; Section VI, Subd. 1, Scholastic Dishonesty.
• University of Minnesota Policy Library: Teaching and Learning: Student Responsibilities — the academic consequence an instructor may impose; excludes the Duluth campus; no visible effective date in the page body.
• University of Minnesota Policy Library: Teaching and Learning: Instructor and Unit Responsibilities — §F, the obligation to report. Effective date and approving body were not visible in the page body as read.
• University of Minnesota Policy Library: Resolving Alleged Student Conduct Code Violations — the advisor, the formal hearing on request, the preponderance standard and the single campus-wide appeal.
• Office for Community Standards: Expungement — third-party release, the seven years, and the AIM process — and what kind of consequences could I face — the outcomes list. Neither page carries a visible date.
• IT@UMN: Turnitin — the two Canvas integrations; no mention of AI detection — and Generative AI Tools: FAQ for Instructors and Students — permission is per course. Page carries no visible date.
• Writing Across the Curriculum: Restricting ChatGPT Use in Classes with Writing Assignments — routes the Turnitin-as-AI-detector question to the Office of Information Technology. Page carries no visible date.
• UMN teaching support: Teaching and generative AI — names the Canvas course “Artificial Intelligence and Academic Integrity: Resources for Instructors”, which requires a UMN log-in via TrainingHub.

This page is orientation, not legal advice. What binds you is the Board of Regents Student Conduct Code, the administrative policies on teaching and learning, your campus conduct procedures, and your course syllabus.