Plagiarism and AI at Ohio State: your instructor must report it

The short answer

At Ohio State your instructor is required to report suspected misconduct. They do not decide the case, and they cannot decide to overlook it. That single rule shapes everything else — including the fact that OSU has enough cases to publish real statistics about them.

It also means the common instinct — have a quiet word, agree a redo — is not available. Understanding that early is worth more than any advice about scores.

Your instructor has no discretion

The rule is short and it is a Faculty Rule, not a guideline:

“Instructors shall report all instances of alleged misconduct to the committee.”

And Ohio State restates it for students:

“By university rule, an instructor must report any suspected instance of academic misconduct to the Committee on Academic Misconduct (COAM). A review panel of the committee will investigate the charges, decide whether or not a violation has occurred, and if the panel finds there has been an offense, determine an appropriate penalty.”

This cuts both ways, and the second way is genuinely in your favour. Yes, a suspicion becomes a formal process rather than a conversation. But it also means the person who suspects you is not the person who judges you, cannot impose a penalty on their own, and has to put the allegation in front of a committee that includes students.

Practically: if an instructor raises a concern, treat it as the start of something formal from that moment. Gather your drafts and version history straight away rather than after the notification arrives. Our guides on proving you wrote it yourself and what to do when you are accused cover what to assemble.

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.

Who actually decides

The Committee on Academic Misconduct is a standing committee of the University Senate, and its composition is unusual:

“The committee on academic misconduct shall consist of sixty-two members. Thirty faculty, at least half of whom shall be tenure-track faculty. … Thirty students. Fifteen graduate students. Fifteen undergraduate students. … Two administrators.”

Students hold half the seats. A panel hearing needs “at least three members of COAM and requires at least two faculty representatives and one student representative”.

Since October 2024 you can choose your forum: a panel hearing, or an administrative hearing before a single member. And “if a student takes responsibility for allegations of academic misconduct and waives their right to a hearing, they may have the allegations resolved as an administrative decision”. Note what that waiver costs — see the appeal section below.

The standard of proof is stated in the Code: “A student will only be found in violation if a preponderance of evidence supports the charges.” More likely than not — not certainty.

The numbers Ohio State publishes

Almost no university publishes this, and Ohio State has done so since 2000-01. From its most recent annual report, covering cases resolved between 6 May 2024 and 4 May 2025:

• 1,148 cases resolved. 70.6% as an administrative decision, 29.4% by hearing.
• 1,088 ended in a finding of “in violation” — 94.6%. Thirty-seven were not in violation; 23 were invalidated, for reasons including that “weight of evidence did not support the allegation”.
• 292 charges of unauthorised generative AI use — 26.0% of cases — of which 93.5% resulted in a finding.

For comparison within the same year: plagiarism charges numbered 353, falsification 375, and violation of course rules 1,090 — the last appearing in nearly nine cases in ten, because it is the catch-all that accompanies most other charges.

Three things to hold on to when reading those figures:

• These are charges, not students. One case usually carries several charges, which is why the percentages sum to well over 100.
• They are counts, not a risk rate. Ohio State enrols tens of thousands of students; 1,148 cases cannot be turned into your personal odds, and anyone who does that is inventing a number.
• The 94.6% finding rate is not a conviction rate for the innocent. Cases are filtered before they get there — 23 were invalidated outright — and most resolve administratively because the student accepts responsibility.

What the figures genuinely show is that AI allegations are now a substantial and ordinary part of the caseload, not an exotic event. The report attributes rising charge counts to “generative artificial intelligence allegations” — without naming any detection tool.

AI is in the Code itself

Many universities handle AI through guidance layered on top of an older rule. Ohio State put it in the Code, approved by the Board of Trustees on 16 November 2023, as its own numbered category:

“Unauthorized use of generative artificial intelligence systems or similar technologies to complete academic activities.”

The default is set out plainly:

“These generative AI tools should not be used in the completion of course assignments unless an instructor for a given course specifically authorizes their use. Some instructors may approve of using generative AI tools in the academic setting for specific goals. However, these tools should be used only with the explicit and clear permission of each individual instructor, and then only in the ways allowed by the instructor.”

Prohibited unless permitted — and permission is per instructor and per purpose. “My other professor allowed it” is not a defence.

Two neighbouring categories catch people out. Reusing your own work without permission is listed, and expressly “includes submitting the same work for courses that the student is retaking pursuant to the university’s grade forgiveness rule”. And the first item on the list is simply “violation of course rules and/or assignment guidelines as contained in the course syllabus” — which is why it appears in almost every case. At Ohio State, breaking the syllabus rule is the offence.

Ohio State does not publish a prescribed disclosure form; the permitted scope and any acknowledgement come from your instructor. Our AI disclosure statement guide covers what a good one contains.

What OSU says about similarity scores

Ohio State runs Turnitin inside CarmenCanvas, branded internally as SimilarityCheck. On whether the AI-writing indicator is switched on, it publishes nothing either way — its two dedicated Turnitin pages list every assignment setting an instructor can configure and never mention an AI score. That is a documented absence, and we will not guess at it.

What OSU does say about the similarity number is unusually candid, and useful if you are ever shown one:

“Instructors should note that there is no Similarity Report outcome that unequivocally indicates plagiarism. The percentage score generated by Turnitin can be influenced by instructor choice in setting up the assignment.”

It then publishes explicit false-positive and false-negative lists. Among the causes of a high score with no wrongdoing: “A student has independently submitted his or her paper to Turnitin.com to check for plagiarism”; “Your student submitted a draft of his paper earlier in the term and it was added to the Ohio State database”; and “The student’s original work is being plagiarized by another student. A Similarity Report does not identify which text is the original and which text is the copy.”

Note the first of those. Running your own paper through a checking service can be the reason it later matches. That is a real and under-appreciated risk with any tool that stores what you give it — ours does not add your text to a third-party comparison database, but check before you use anything.

There is also a fairness instruction worth knowing: “If you plan to run SimilarityCheck on an assignment, we strongly recommend that SimilarityCheck be run for all students in the course as opposed to a select few.” If you have been singled out, that recommendation exists.

More on what these numbers can and cannot support in what a Turnitin score actually means and how accurate AI detectors really are. The same limit applies to our own tool: our published error rates are measured on English and German corpora, and no detector produces proof.

Your paper joins a database

Ohio State’s sample syllabus statement says so directly:

“Note that submitted final papers become part of the Ohio State database.”

And the opt-out belongs to your instructor, not to you: “Faculty may opt to not include papers in the Ohio State Turnitin database.” There is no published student-side opt-out and no stated retention period — which is a meaningful contrast with Toronto and Waterloo, where students can decline outright.

On self-checking: “Turnitin is licensed explicitly for classroom or teaching use only. Faculty and graduate students wanting to check their own work should use iThenticate.” Undergraduates have no self-service route. What OSU does recommend to instructors is worth asking for: “Create separate Assignments for draft submissions and final submissions”, and “Make reports available to students.”

Sanctions and the record

The Code lists formal reprimand, disciplinary probation, suspension and dismissal — the last “permanently separates a student from the university without opportunity to re-enroll in the future”. Separately, “a hearing body may authorize the instructor to award a failing or lowered grade in the course and a loss of credit on the graded coursework”.

There is no transcript notationin the Code — no equivalent of Penn State’s XF or UCF’s Z. But there is a record, and Ohio State is honest about where it surfaces:

“A record of the finding will be retained in the Office of Student Conduct, and you may need to explain it to any employers or graduate and professional schools that require, as part of their application, formal confirmation that you do not have a record of academic misconduct.”

Whether such a record can ever be removed is not published. Sanctions “may be more severe for students previously found in violation”, and the hearing body weighs mitigating circumstances, past misconduct, actual and potential harm, “the degree of intent and motivation”, and how pervasive the conduct was.

Five business days, and one risk

The appeal window is the shortest in this series: five business days after the decision is sent. One appeal only, and “the decision of the appeal officer is final”. Appeals from COAM go to the executive vice president and provost or designee.

Grounds are limited to three:

• Procedural error causing material harm or prejudice.
• Substantial new evidence unavailable at the time of the hearing.
• A sanction “grossly disproportionate to the violation(s) committed”.

Accepting responsibility narrows this to almost nothing: you “waive the right to appeal, except on the basis that the disciplinary sanction is grossly disproportionate”. If you intend to contest the finding, do not take the quick administrative resolution.

And there is a genuine downside risk on appeal:

“If a case is reheard by a hearing body, the sanction imposed can be greater than that imposed at the original hearing.”

Appealing is not free. Weigh it, rather than filing reflexively.

One helpful change: since October 2025, students “will receive the incident summary — a recap of the allegation narrative submitted by the reporter — as part of their initial notification”. You should learn what you are accused of at the outset rather than at the hearing.

Before you submit

• Read the syllabus rule. Breaking it is itself the first-listed offence, and it appears in nearly nine cases in ten.
• Get AI permission from that instructor, in writing. Permission does not travel between courses.
• Do not reuse your own work, including in a course you are retaking under grade forgiveness.
• Keep drafts and version history. The process is formal from the first suspicion.
• Do not accept responsibility to make it go away if you intend to dispute the finding — it waives your appeal.
• Diarise five business days from any decision.

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.

Sources

• Ohio State: Code of Student Conduct 3335-23-04, Prohibited conduct — Board approved 16 November 2023; AI is a numbered category.
• University Senate: COAM Annual Report, Summer 2024 – Spring 2025 — cases resolved 6 May 2024 to 4 May 2025.
• Office of Academic Affairs: Academic integrity: information for students — page carries no date; banner dated October 2025.
• Teaching and Learning Resource Center: Turnitin best practices — the false-positive and false-negative lists; page carries no date.

This page is orientation, not legal advice. What binds you is your course syllabus and the Ohio State Code of Student Conduct.