Plagiarism and AI at UT Austin: the rules never mention AI, so your syllabus decides

The short answer

UT Austin’s academic misconduct rule does not mention artificial intelligence anywhere — so what governs your AI use is your syllabus, read through the rule’s existing catch-alls. Get the AI policy for each course in writing, because that is the document a case will actually turn on.

The second thing to know is procedural, and it costs students their appeal more often than any rule about AI: if you sign an Administrative Disposition agreeing to both the finding and the sanction, the matter is final and cannot be appealed at all.

The rule has no AI clause

The operative instrument is Chapter 11 of the Institutional Rules on Student Services and Activities. Its own closing line dates it: “Updated September 2025 to reflect policy changes effective September 1, 2025.” Above it sit the UT System Board of Regents’ Rules — Chapter 11 says it operates “in accordance with The University of Texas System Rules and Board of Regents’ Rule”.

Sec. 11-401 defines academic misconduct as “any of the following activities”, and this is the complete list:

“Copying– copying answers or information from another individual’s academic assignment; Failure to Comply with Instructions– failing to comply with instructions pertaining to an academic assignment or course requirement, and where that failure results in an unfair academic advantage; Unauthorized Materials– obtaining, accessing, or possessing materials without authorization, including but not limited to class notes, textbooks, calculators, online resources, and/or electronic devices; Unauthorized Aid or Assistance – providing aid or assistance to, or utilizing aid or assistance from, another individual or source without authorization, and pertaining to an academic assignment or course requirement; Substitution; Falsification or Fabrication – falsifying or fabricating any information, data, or citation in any academic work offered for credit; Plagiarism– when a person represents another’s material as their own work without attribution; when a person misrepresents citation or attribution for purposes of an academic advantage; or, when a person submits essentially the same work for two assignments without the permission of the Faculty Member. Collusion– unauthorized collaboration with another student or students; Failure to Follow Course Requirements; Other Academic Misconduct– any activity in which a student receives or attempts to receive an unfair academic advantage.”

Ten categories, and no artificial intelligence. A full-text search of Chapter 11’s rule body for “artificial”, “generative”, “AI” and “ChatGPT” turns up exactly one hit, and it is an “artificial body of water” in Sec. 11-402, on fountains.

That is a real difference from other large US universities. Ohio State wrote AI into its Code as a numbered clause; Texas A&M names it inside the definitions of Cheating and Fabrication. UT Austin, updating its rules in September 2025, did not.

This is not good news, and it should not be read as a gap you can stand in. It means AI is prosecuted through the existing categories — most obviously Unauthorized Aid or Assistance(“utilizing aid or assistance from, another individual or source without authorization”), Failure to Comply with Instructions, and the residual Other Academic Misconduct, which reaches “any activity in which a student receives or attempts to receive an unfair academic advantage”.

The Center for Teaching and Learning states the link explicitly: work “generated by an AI system that is not credited to that system” violates existing academic honesty policies, and using AI “without permission or authorization, or failing to properly cite AI” constitutes “a violation of UT Austin’s Institutional Rules on academic integrity.”

And the Dean of Students warns that meaning to do it is not part of the test:

“Academic misconduct, commonly referred to as ‘cheating,’ undermines learning and threatens the university responsibility to certify student progress and educational achievement. These violations, including plagiarism, copying and collusion, among others, can have serious consequences, even if the misconduct was unintentional.”

So the syllabus decides — three models

Because the rule is silent, UT’s institutional guidance is a menu that instructors choose from. The Center for Teaching and Learning publishes three sample syllabus positions. The page is undated, and it is now served from fisd.utexas.edu; the Dean of Students points faculty to it directly (“Consider adding a statement about ChatGPT and AI to your syllabus. Visit the Center for Teaching and Learning’s 5 Things to Know about ChatGPT to learn more.”).

1. No AI use permitted.“This course assumes that work submitted by students will be generated by the students themselves” — and, bluntly, “the use of generative AI tools for any part of your work will be treated as plagiarism.”
2. AI permitted with acknowledgment.“students shall give credit to AI tools whenever used, even if only to generate ideas”; “it is a violation of university policy to misrepresent work by not acknowledging generative AI tools”.
3. AI use encouraged.“I expect you to clearly attribute what text was generated by the AI tool”, and “you are required to include a paragraph that explains what AI content-generation tool you used”.

Notice how far model 2 reaches: credit is owed even where the tool only helped you generate ideas, not sentences. That is a wider duty than most students assume, and it is the model most likely to catch someone acting in good faith.

Where disclosure is required, the guidance asks for four elements: “(a) the entire exchange, (b) which AI tools were used, (c)” how they were used, and “(d) why AI tools were used.” The first of those is unusual — it contemplates handing over the conversation itself, not a summary of it. If your course uses that model, keep your transcripts.

On citation format, the page “directs instructors to consult MLA, APA, and Chicago style guides for AI citation formats but provides no specific internal citation instructions”. So there is no UT house form. Our guide to writing an AI disclosure statement covers the general shape; at UT, match your syllabus and your style guide.

A university-wide required disclosure wording does not exist as far as we could establish. There is no single form of words that keeps you safe everywhere on campus — which is exactly why the per-course email matters.

The sanctions, and the transcript

Academic sanctions, Sec. 11-702:

“When a student is found responsible for academic misconduct, as defined in Sec. 11-401 … Student Conduct or the Dean of Students, may impose any of the following sanctions: Academic integrity probation; no credit or reduced credit for the graded assessment in question; retaking of examination or resubmission of assignment; failing or reduced final grade for the course in question.”

Disciplinary sanctions under Sec. 11-701(a) can be added on top, and the published list runs: “written warning; disciplinary probation; educational sanction; restorative practice; bar against re-admission; bar against enrollment; withdrawal from one or more classes; restitution … suspension of rights and privileges … deferred suspension; suspension from the University for a specified period of time; expulsion (permanent separation from the University); other sanction or sanctions as deemed appropriate under the circumstances.”

There is also a programme-removal rule for competitive programmes — those requiring a supplemental application, a minimum GPA to matriculate, and a signed commitment to a programme honour code: for those, “removal from the program will be considered by Student Conduct as a potential sanction.”

On the transcript, UT is different from the universities that brand the grade. There is no F*, no XF, no dishonesty-specific notation of any kind. Notations attach only to separation, under Sec. 11-704:

“Suspension is noted on the student’s official transcript in accordance with Sections 9-301 and 9-402 of the Institutional Rules.” — “A permanent notation of expulsionwill also be placed on the student’s official transcript”. — “A bar from readmission or reenrollment will be noted on the student’s official transcript”.

So a failing grade for academic misconduct at UT looks, on the transcript, like any other failing grade. The expulsion notation is stated to be permanent. Whether a suspension notation can ever be removed is governed by Sections 9-301 and 9-402, which we were not able to retrieve — so we cannot tell you, and you should ask Student Conduct directly rather than assume either way.

Two further powers are worth knowing about, because they bite at the end of a degree:

“Rescission of Degree– The University may rescind a degree already conferred where it has been determined … that the degree holder, while enrolled as a student and while in the process of completing their degree requirements, engaged in academic misconduct … and where a finding of that violation also affects whether the degree holder completed the applicable degree requirements.” (Sec. 11-704)

“Student Conduct may withhold an official transcript, grade, diploma, or degreeof a student found to have violated a rule … Student Conduct may take such action pending a hearing, resolution by Administrative Disposition, and/or exhaustion of appellate rights” where it judges that the best interests of the University or the UT System would be served. (Sec. 11-703)

And a suspension for academic misconduct specifically blocks the usual workaround of taking the credit elsewhere: the Registrar and your department may permit credit for work done at another institution during a suspension, “except in cases where suspension is imposed for academic misconduct.”

Your instructor does not decide

This is structurally important and it works in your favour. At UT, faculty who want to pursue a case must hand it over:

“To resolve potential academic misconduct, faculty and instructional staff … are encouraged to first engage with the student involved. These conversations often provide valuable insight into the facts and the student’s perspective. In some cases, faculty may discover insufficient evidence to support the allegation or learn that the issue stems from a misunderstanding rather than intentional misconduct.”

“If faculty and instructional staff determine they would like to pursue an academic misconduct case, they must refer the alleged violation to Student Conduct and Academic Integrity (SCAI)in the Office of the Dean of Students through the online Academic Misconduct Report Form.”

There is a filing deadline on the instructor: “Submit your referral no more than 30 daysafter you become aware of or should have reasonably become aware of the alleged violation.” And referrals can fail for thinness: “Referrals that lack the necessary documentation may be dismissed without further investigation.”

The rules on what happens to you meanwhile are explicit:

• “Students engaged in the academic misconduct process must be permitted to complete all assignments and attend classesuntil the matter is resolved.”
• “Students should be assigned an Incomplete (‘X’) grade until the conduct process is complete and both the student and instructor have received the outc me document from Student Conduct.” (The missing letter is in the original.)
• “Students may not Q-drop a courseonce a faculty member or Student Conduct notifies them of suspected academic misconduct, during the conduct process, and/or if the student has been found responsible”.

Then Student Conduct investigates, under Sec. 11-501: “Student Conduct will perform an initial assessment and determine … if an investigation is merited, or if the referral shall be maintained without further action. If further investigation is merited, Student Conduct will investigate the allegation(s) and afford the accused student an opportunity to hear and respond.”

The investigation ends one of two ways: a No Finding Letter “in instances where the preponderance of the evidence does not support the finding of a violation”, or an Administrative Disposition setting out “the results of the investigation; the basis for the finding of a violation; the specific section(s) of this Chapter the student has violated; and any assessed sanction(s).”

If it goes to a hearing, the deciding structure is layered: a Student Conduct Panel of Hearing Officers and Student Conduct Board members “conduct hearings and issue a written responsibility recommendation to the Dean of Students”, and then “the Dean of Students will render a written Disciplinary Decision, which includes the Responsibility Decision and Sanction(s). … The written decision is the official outcome on the matter from which any appeal is based.”

The Dean of Students’ own advice to a suspected student is short and practical:

“The student conduct process prioritizes due process. You will receive instructions from your instructor or Student Conduct staff. Keep any documents, emails, chats, or other evidence that could help you if you believe the suspicion is in error. Student Conduct staff will offer to meet with you to hear your side and review the evidence. They will also fully explain the conduct process and your rights.”

Notice that this only works if the material still exists. Our guides on proving you wrote it yourself and what to do when you are accused are about building that record before you need it.

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.

Preponderance, stated three times

UT states its standard in the rule, in the hearing provisions, and again in its instructions to faculty — which is a good sign for consistency.

Sec. 11-102(d): “University disciplinary action instituted by Student Conduct will be based upon the ‘Preponderance of Evidence’ standard. This standard is satisfied if the alleged violation is deemed more likely to have occurred than not.”

Sec. 11-608: “All university hearings shall use the ‘preponderance of the evidence’ standard … Student Conduct has the burden of providing evidence and proving the chargesby the preponderance of the evidence.”

And to faculty:

“This standard of information requires supporting documentation that demonstrates it is more likely than not that the violation occurred. All supporting documentation must be shared with the studentand SCAI.”

That last sentence is one to hold onto. If a case is being made against you, the documentation behind it is supposed to reach you as well as the office. Ask for it in those terms.

The signature that ends your appeal

UT’s appeal grounds are the familiar three, from Sec. 11-801(d):

• “Significant procedural error inconsistent with the processes as outlined in subchapters 11-500 and 11-600.”
• “Discovery of any new information unknown or not reasonably foreseeable to the accused student or Student Conduct at the time of the hearing that was material to and could have reasonably impacted the Disciplinary Decision.”
• “The sanction(s) determined by Student Conduct or the Dean of Students are significantly disproportionate to the violation.”

The deadline is five days, and the decision of the Appellate Officer “shall be final and will conclude the appellate process.” Unusually, the appeal right runs both ways: “Either Student Conduct or the student may request to appeal a hearing outcome to the Appellate Officer.”

And here is the trap. Sec. 11-802(a):

“An Administrative Disposition, presented in accordance with Sec. 11-501(b)(2), and in which the student agrees to both the finding of responsibility and the sanction(s) assessed by Student Conduct, is final and is not eligible for appeal.”

An Administrative Disposition is offered as the quick way out of a stressful process, and for a lot of students it is the right choice. But agreeing to both halves of it — the finding and the sanction — closes the door completely. There is no reopening it later because the consequence turned out to be worse than you expected.

The partial route keeps one narrow option alive:

“A student who accepts a finding of responsibility … but chooses to contest only the sanction(s)assessed … may submit a written appeal … The written appeal may only cite ‘Disproportionate Sanctions’ … as the basis for the appeal.”

And that appeal is decided on paper alone: “The appeal of the sanction(s) … will be reviewed solely on the basis of the Administrative Disposition, the student’s written appeal submission, and the response to the appeal submitted by Student Conduct, if any. Oral argument will not be considered.”

So before you sign anything, understand which of the two boxes you are agreeing to. Accepting the finding but contesting the sanction is a real, published option with a narrower door still open behind it.

Filing on time does buy protection, under Sec. 11-804: “A timely appeal suspends the imposition of the sanction(s) until the appeal is final” — though interim actions stay in place, and transcripts, diplomas, grades or degrees “may also be withheld pending conclusion of the appeal”.

What UT tells staff about tools

The Office of the Dean of Students names two products to faculty, on an undated page, under a “Tools” heading:

“For papers, use TurnItIn or iThenticate to check for plagiarism.” “Canvas has options to check for and prevent academic misconduct.” “When creating assessments in Canvas, use question groups/banks so students will see different questions in different orders.”

Immediately after that list comes an instruction that is worth quoting to anyone who flags you:

“For all these tools, it is important that if a tool flags a student for further review, the information must be reviewed to see if there has been a potential violation or a false flag.”

A flag is a prompt to look, in UT’s own instruction to its staff — not a conclusion.

On the specific question everyone asks — is a Turnitin AI indicator switched on at UT? — we found no statement either way, on the Dean of Students conduct pages or on the CTL’s AI pages. The nearest thing is a report of a single test result on the CTL’s ChatGPT page, and it is about the similarity checker, not an AI indicator:

“A ChatGPT-generated paper did not flag the Turnitin plagiarism detector tool (i.e., no plagiarism was detected).”

That says a similarity checker does not catch generated text, which is unsurprising: a similarity checker looks for matching sources, and generated text usually matches none. It is not a statement about AI detection, and we will not stretch it into one. We do not assert that any detector is accurate or inaccurate — including ours. Our published error rates are measured on English and German corpora, and no detector produces proof; see how accurate AI detectors really are and what a Turnitin score actually means.

For students, there is no self-check route documented. The Turnitin and iThenticate instruction is addressed to faculty, and UT’s own student-facing advice is human rather than technical:

“Get help if you think you might need it. Sanger Learning Center and University Writing Center offer free support.” “Check your work. Give yourself enough time to thoroughly review each assignment and be sure to credit the sources of any borrowed material properly, using the citation style specified by your instructor.” “Assume all work is to be done individually unless otherwise stated by your instructor” “Perceived ‘shortcuts’ for an exam or assignment are often forms of academic misconduct.”

What UT does not publish

Being straight about the edges of what we could establish, because guessing here would be worse than useless:

• Whether an AI indicator is enabled in Turnitin. No statement in either direction was found on the pages we could read.
• Whether your submissions join a comparison repository, for how long, and whether you can opt out. Nothing on this is published on the Dean of Students or CTL pages. The operative Canvas instructions for setting up a Turnitin assignment sit behind the UT login, so that document is unreadable to us, not absent — a UT student or staff member can open it.
• Case statistics.No annual report or case-count page exists in the Dean of Students’ conduct section, so nobody outside UT can say how many AI allegations there have been.
• Sections 9-301 and 9-402, which govern transcript notations, were not retrieved.

Two retrieval notes, so you can find things yourself. The CTL’s AI pages now redirect from ctl.utexas.edu to fisd.utexas.edu, which refuses plain automated requests — they open normally in a browser, and the quotes above are short because our reading of them was length-limited, not because the pages are thin. And academicintegrity.utexas.edu did not resolve at all when we looked; the academic-integrity function lives under the Office of the Dean of Students.

What isdocumented is the confidentiality of the disciplinary record itself, Sec. 11-901: “The Disciplinary Record will be separate from the student’s academic record, will be confidential, and will not be revealed except on request of the student or in accordance with applicable state and federal law, or as otherwise provided in subchapter 9-300.”

Before you submit

• Email your instructor for the AI policy and keep the reply. The rule is silent, so the syllabus is the instrument — and an email is evidence of what you were told.
• If your course uses the “with acknowledgment” model, credit the tool even for ideas.That model asks for credit “even if only to generate ideas”.
• Keep the whole exchange. Where disclosure is required, one of the four elements asked for is the entire exchange, not a summary.
• Check every citation you did not fetch yourself. Falsification or Fabrication covers fabricating “any information, data, or citation”.
• Do not sign an Administrative Disposition until you understand it. Agreeing to the finding and the sanction ends the matter with no appeal; agreeing to the finding alone leaves a narrow sanction appeal open.
• Do not try to Q-drop. That route closes the moment you are notified.

If you want a private read on your own writing before you hand it in, our AI check is free up to 1,500 words, with no account and no name. It is a second opinion, not a clearance certificate.

Sources

• The University of Texas at Austin: Institutional Rules on Student Services and Activities, Chapter 11 — Student Conduct and Academic Integrity — “Updated September 2025 to reflect policy changes effective September 1, 2025”; operates under the UT System Board of Regents’ Rules and relates to Handbook of Operating Procedures 3-3031.
• Office of the Dean of Students: Resolving Academic Misconduct — undated; the referral duty, the 30-day deadline and the preponderance statement to faculty.
• Office of the Dean of Students: Avoiding Academic Misconduct — undated; the student-facing advice and “If You Are Suspected of a Violation”.
• Office of the Dean of Students: Academic Integrity and Classroom Expectations — undated; the “Tools” list and the false-flag instruction.
• Center for Teaching and Learning: 5 Things to Know about ChatGPT — undated; served via a redirect from ctl.utexas.edu, and it refuses automated requests, so read it in a browser.
• Center for Teaching and Learning: ChatGPT and Generative AI Tools: Sample Syllabus Policy Statements — undated; the three instructor-chosen models.

This page is orientation, not legal advice. What binds you is your course syllabus, Chapter 11 of the Institutional Rules on Student Services and Activities, and the UT System Board of Regents’ Rules above it.