Academic integrity and AI at Rutgers: the XF grade, 18 months, and why a thesis is Level 3

The short answer

Rutgers can put an X in front of a failing grade, and the X comes off only by written petition — which will not even be considered until at least 18 months after the violation. That is the consequence to understand first, because it outlasts the course by years.

The second fact matters if you are writing a thesis. Policy 10.2.13 names “a capstone project, thesis, or dissertation” at Level 3, the most serious tier — the only document type the policy singles out anywhere. The same act on a thesis is classified more severely than on a term paper.

The XF grade and the 18-month wait

Rutgers publishes two transcript consequences. Expulsion carries “a permanent notation of disciplinary expulsion on the student’s transcript”. The XF is the other, and it is the only one with a published route off. Here it is in full, because every clause of it is operative:

“Requests for removal of the X from an XF grade must be submitted in writing to the appropriate Campus Appeals Committee. Such requests will not be considered until at least 18 months from the time of the violation that resulted in the XF. In order for the request to be granted, the student, at a minimum, following the original violation, must have an exemplary record with respect to academic integrity, must have completed a Rutgers academic integrity workshop or seminar, and must satisfactorily answer a required series of essay questions on why the X should be removed. The Campus Appeals Committee shall make the final decision concerning the request. If the request is denied, the student must wait another year to submit another request.”

Unpacked, that is:

• Minimum 18 months before anyone will even look at your request.
• Four conditions, all of them: written request, an exemplary integrity record since, a completed Rutgers workshop or seminar, and satisfactory essay answers.
• A denial costs a further year. So a weak first attempt is expensive.
• The Campus Appeals Committee decides, and its decision is final.

Note also that the XF appears in the policy as “A grade of XF in the course, where applicable” — it is one option among the academic penalties, not automatic. The others are resubmission or an alternate assignment, a grade reduction for the assignment or the course, a failing grade for the assignment, and a plain F in the course.

Why a thesis sits at Level 3

If you take one thing from this page as a thesis writer, take this. The Level 3 examples in Policy 10.2.13 are, verbatim:

“Any violation involving potentially criminal activity.”; “Coordinating an organized cheating scheme.”; “Having a substitute take an examination.”; “Cheating and/or plagiarism on a capstone project, thesis, or dissertation.”; “Intentionally destroying another student’s work.”; “Falsifying patient or client records.”

And Level 3 is defined as: “Level 3 violations represent the most serious breaches of conduct. They may involve a serious violation of a professional code of conduct; may include extreme cases of dishonesty and maliciousness or violations of law; and/or are likely to cause direct harm to others.”

So the company your thesis keeps in that list is impersonation at an exam and falsifying patient records. Not because Rutgers thinks a citation slip is a crime, but because the document carries the degree.

What follows practically: the margin for sloppiness that exists on a weekly assignment does not exist on your thesis. A Level 1 violation is described as one that “may occur as a result of inexperience or lack of malicious intent” and is “generally quite limited in extent”. That framing is not available to you on a dissertation chapter. Getting quotation marks, paraphrase distance, source verification and any AI use exactly right on the thesis is worth more care than anywhere else in your degree.

There is also an interim measure attached to Level 3 allegations. Where the allegation involves law or a professional code, or direct harm to others is likely, the student “may be removed from a course, clinical, or internship setting on an interim basis, with the approval of the dean of the school”. The student must be “immediately informed in writing” and “has the right to appeal to the campus Chief Academic Integrity Officer”.

The three levels, and how they are assigned

The policy publishes three levels, not four — worth saying because summaries elsewhere sometimes say otherwise.

“Violations of academic integrity are generally divided into three categories: Level 1, Level 2, and Level 3.”

• Level 1 — “may occur as a result of inexperience or lack of malicious intent by the person committing the violation.” Examples include plagiarism on a minor assignment or a very limited portion of a major one, unpremeditated cheating on a quiz, prohibited collaboration on homework, unauthorized sharing of course materials, “Citing a source that does not exist or that one has not read on a minor assignment”, and signing in for another student.
• Level 2 — “misconduct of a more serious character or misconduct that affects a major, significant, or essential portion of work done to meet course requirements. These violations demonstrate premeditation or may have posed harm to others.” Examples: “A second violation.”; “Substantial plagiarism on a major assignment.”; unauthorized materials or collaboration on a major exam; falsifying data for a major assignment; “Distribution of course materials for financial gain.”; facilitating dishonesty by another student; obstructing another student’s work; participating in an organized cheating scheme.
• Level 3 — as above.

Two of those deserve a flag. “A second violation” is itself a Level 2 example, so a small first finding raises the floor for anything that follows. And “Citing a source that does not exist or that one has not read” is named at Level 1 for a minor assignment — which is precisely the failure mode of an unverified AI-generated bibliography. On a thesis, the same fabricated citation is not a minor-assignment matter. Check every reference: how to verify your citations is the checklist.

Classification is not the instructor’s alone. It is made “by the Academic Integrity Facilitator (who is appointed in each School to investigate and adjudicate alleged academic integrity) in consultation with the campus Office of Student Conduct or equivalent.”

The policy, its date, and where AI lands

The governing document is the Academic Integrity Policy, Section 10.2.13. Its approval authority is the Senior Vice President for Academic Affairs. It was adopted 2 September 2008, reviewed 2 June 2020, and revised on 1 April 2010, 1 September 2011, 1 September 2013 and 2 June 2020. Its scope is stated at the top: “This Academic Integrity Policy applies to all schools and academic units of Rutgers, The State University of New Jersey.” University-wide, not campus-specific.

Its own footer, on every page, is a caution worth repeating: “All policies are subject to amendment. Please refer to the Rutgers University Policy Library website (policies.rutgers.edu) for the official, most recent version.” Check there before relying on any date, including ours.

The pledge:

“On my honor, I have neither received nor given any unauthorized assistance on this examination (assignment).”

The words “artificial intelligence”, “generative AI” and “ChatGPT” do not appear anywhere in Policy 10.2.13. The last revision was June 2020 — before ChatGPT existed — which explains the silence and is worth stating plainly: the levelled table above was not written with generative AI in view.

The hook an AI case runs on is the first principle of academic integrity, verbatim: students must

“make sure that all work submitted in a course, academic research, or other activity is the student’s own and created without the aid of impermissible technologies, materials, or collaborations.”

And, under Cheating: “Using or having access to books, notes, calculators, cell phones, technology, or other prohibited devices or materials during a quiz or examination.”

Impermissible and prohibited are the load-bearing words, and neither is defined by the policy. They are defined by your course.

The other definitions you should know, verbatim:

• Plagiarism— “the use of another person’s words, ideas, images, or results, no matter the form or media, without giving that person appropriate credit.”
• Cheating— “the use or possession of inappropriate or prohibited materials, information, sources, or aids in any academic exercise. Cheating also includes submitting papers, research results or reports, analyses, and other textual or visual material and media as one’s own work when others prepared them.”
• Fabrication — “the invention or falsification of sources, citations, data, or results, and recording or reporting them in any academic exercise.”
• Facilitation of Dishonesty— “deliberately or carelessly allowing one’s work to be used by other students without prior approval of the instructor or otherwise aiding others in committing violations of academic integrity.” Note carelessly: lending a friend your chapter “just to see the structure” is inside this definition.
• Academic Sabotage — “deliberately impeding the academic progress of others.”
• Violations Involving Potentially Criminal Activity — “theft, fraud, forgery, or distribution of illicitly obtained materials committed as part of an act of academic dishonesty.”

Turnitin at Rutgers is instructors only

This one is stated flatly by Rutgers University Libraries, in an Ask a Librarian answer last updated 22 January 2024:

“Rutgers does subscribe to Turnitin (a plagiarism detection software) but its use is limited to instructors only.”

Turnitin “is integrated into all Rutgers learning management systems, but instructors can also apply for a standalone account if preferred.”

So there is no Rutgers-provided way for you to see your own similarity report before submitting. No Draft Coach, no practice assignment, no writing-centre similarity route is published. The Libraries’ answer directs students to third-party tools instead — which is worth knowing, and worth being careful with, because a third-party tool is somewhere your unpublished thesis now lives. Before you upload a draft anywhere, find out whether the service keeps it.

Turnitin is not named anywhere in Policy 10.2.13 or on the Rutgers academic integrity site. And whether Turnitin’s AI-writing indicator is switched on in the Rutgers tenancy is not publicly documented. No on/off decision is published either way, and silence should not be read as “on”.

What is documented at Rutgers, and it is not nothing, is the repository question — or rather its absence. No retention statement and no opt-out was located.

What the Rutgers Libraries tell instructors

No tenancy setting is published, but Rutgers does publish an argument. The Rutgers University Libraries research guide Artificial Intelligence: AI in Teaching, last updated 1 September 2026, has a section headed “AI Detection Tools”. If you are ever accused on the strength of a score, this is the passage to have:

“While tools for detecting AI-generated content exist (Turnitin, GPTZero, Copyleaks), such tools are not 100% effective and have been found to exhibit several problems. For instance, AI detectors:

Can be prone to produce false positives or negatives

Offer no explanation or supporting evidence for their scores

Can easily be evaded through the manipulation of prompts or rewriting outputs

Tend to be biased against non-native English writers

Do not guarantee privacy of any data submitted”

And what it recommends instead:

“Instead of relying on AI detection tools, it is recommended that instructors: Implement clear policies and guidelines governing the use of AI in the classroom; Teach students how to use AI ethically and effectively and critically evaluate outputs; Create scaffolded assignments that encourage intrinsic motivation and focus more on process than product; Develop rubrics that emphasize critical thinking, problem solving, and applied knowledge rather than the memorization and summarization of content”

Be precise about what this is when you cite it. It is a Rutgers University Libraries research guide — guidance from the Libraries, not a provost or senate policy. It does not bind an instructor and it does not decide a case. What it does is establish that Rutgers’ own librarians, publishing on a Rutgers domain and updated this month, list false positives and bias against non-native English writers among the known problems with these tools. In a meeting where a percentage is the entire evidence, that is a reasonable thing to put on the table.

The two background pages worth reading alongside it are how accurate AI detectors really are and AI detectors and non-native speakers. If a case has already started, what to do when you are accused comes first.

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.

Disclosure: three model syllabus statements

Rutgers has no university-wide AI disclosure requirement. Policy 10.2.13 requires that students “properly acknowledge and cite all use of the ideas, results, images, or words of others” and “properly acknowledge all contributors to a given piece of work”, but it names no AI-specific form.

The Libraries’ guide states the reading of 10.2.13 that Rutgers applies to AI:

“Rutgers’ Academic Integrity Policy (10.2.13) states that students must ensure ‘that all work submitted in a course, academic research, or other activity is the student’s own and created without the aid of impermissible technologies, materials, or collaborations.’ Whether use of a technology is permissible depends on the specific policies and learning objectives of the course.”

It adds: “Instructors should set clear guidelines about the permissability [sic] and appropriate use of A.I. in their courses.” The misspelling is in the original.

It then offers three model syllabus statements. Knowing them is useful, because the one your course adopts is the one you are graded against — and the third tells you what full disclosure looks like at Rutgers:

“Example 1: Use of AI strictly prohibited — Use of AI tools such as ChatGPT to complete course assignments is not permitted under any circumstances.”

“Example 2: Use of AI permitted under certain conditions — Use of AI tools such as ChatGPT is only permitted to help you brainstorm potential research topics, develop research questions, or find more information about a topic. All material you submit must be your own.”

“Example 3: Use of AI fully permitted — Use of AI tools such as ChatGPT is fully permitted provided that you include the exact prompt used to generate the content as well as the AI’s full response in an Appendix. Because AI generated content is not always accurate or appropriate, it is each student’s responsibility to assess the validity and applicability of any AI-generated content included in their work.”

Example 3 is the one to plan for even if your course uses Example 2: if you keep your prompts and the full responses as you go, you can produce the appendix, and you can also produce the record if anyone asks how the text came about. Our AI disclosure statement guide has wording that fits this shape.

Note also that Example 2 permits brainstorming and research questions while requiring that “All material you submit must be your own” — a distinction between thinking help and text generation that is easy to blur at 2am.

Proof, the TZ grade, and not being able to drop

The standard of proof is stated in one sentence:

“The standard of proof in academic integrity cases is by the preponderance of the evidence, which means that the person or hearing panel deciding the case finds that it is more likely than not that the accused student is responsible for the alleged violation.”

More likely than not — decided by the person or panel hearing the case. That is a lower bar than a criminal standard, and it is a reason to bring concrete material rather than argument alone. Drafts, dated files, outlines, notes, your reading, the emails where you asked what was allowed: how to prove you wrote it yourself is the practical version.

Two procedural facts that catch people out:

• You cannot drop the course. “Once a student has been notified of an alleged violation of academic integrity related to a course, the student may not drop the course until the disciplinary process is completed.”
• Your grade is held, not decided. If a grade is due before the process ends, “the accused student shall be given a TZ grade or similar (e.g., Incomplete) until the disciplinary process is completed.” A TZ on your record mid-process is not a finding.

The detailed procedures are not in the policy itself: “Procedures for adjudicating alleged violations of academic integrity are described at academicintegrity.rutgers.edu.” The named documents are “Procedures for Adjudicating Alleged Academic Integrity Violations”, a separate set for Rutgers Biomedical and Health Sciences schools, and “Academic Integrity Education Recommendations”. Which campus route applies to you is also published — New Brunswick, Camden and Newark each have their own office.

Rutgers publishes no case counts or outcome data that we could locate.

The full sanction range

The policy is explicit that its lists are not exhaustive:

“Any violation of academic integrity is a serious offense subject to appropriate sanctions or penalties. Violations of academic integrity may result in academic penalties, educational sanctions, and/or disciplinary sanctions, and are not limited to the examples below.”

Academic penalties: resubmission or an alternate assignment; a grade reduction for the assignment or the course; a failing grade for the assignment; an F in the course; an XF in the course, where applicable.

Educational sanctions: an assigned paper or research project on ethics or academic integrity; participation in a workshop or seminar; service to the University community.

Disciplinary sanctionsrun considerably further, and several of them are aimed squarely at graduate students: disciplinary reprimand; disciplinary probation; “Dismissal from a departmental or school program.”; “Denial of access to internships or research programs.”; “Loss of appointment to academic positions.”; “Loss of departmental/graduate program/school endorsements for internal and external fellowship support and employment opportunities.”; “Removal of fellowship or assistantship support.”; “Suspension for one or more semesters.”; “Requiring a delay in the completion of a professional program.”; “Dismissal from a graduate or professional program.”; “Expulsion from the University with a permanent notation of disciplinary expulsion on the student’s transcript.”; “Degree or certificate revocation.”

And the policy’s own footnote: “Note: Any of these sanctions may have an impact on financial aid, scholarship, or fellowship eligibility, or scholastic standing.”

For a funded graduate student, losing assistantship support or a fellowship endorsement is often the practically decisive consequence — more than the grade.

Before you submit

• Treat the thesis as the highest-stakes document in your degree. It is the only one Policy 10.2.13 names, and it is named at Level 3.
• Find out which of the three model AI statements your course uses — prohibited, conditional, or fully permitted with prompts and responses in an appendix — and keep the answer in writing.
• Keep prompts and full responses as you go if you use AI at all. You cannot reconstruct them later, and Example 3 is the published disclosure shape.
• Verify every citation. Citing a source that does not exist is named in the policy as a violation in its own right.
• Be careful where you upload drafts for a self-check. Rutgers gives students no Turnitin route, so anything you use is a third party holding your unpublished thesis.
• If you are notified of an allegation, do not try to drop the course. You cannot, and the attempt tells a story you do not want told.

If you want a read on your own text before you hand it in, our AI checkis 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. Rutgers’ own Libraries make the same point to instructors.

Sources

• Rutgers: Academic Integrity Policy, Section 10.2.13 — Approval Authority: Senior Vice President for Academic Affairs; adopted 2 September 2008; reviewed 2 June 2020; revised 1 April 2010, 1 September 2011, 1 September 2013 and 2 June 2020. Applies to all schools and academic units.
• Rutgers: Academic Integrity at Rutgers — links the policy and the three campus routes (New Brunswick, Camden, Newark).
• Rutgers Academic Affairs: Academic integrity policy and procedures — the adjudication procedure documents named by the policy.
• Rutgers University Libraries: Ask a Librarian — does Rutgers use Turnitin? — answered by Ann Marie Latini; last updated 22 January 2024; “limited to instructors only”.
• Rutgers University Libraries: Artificial Intelligence: AI in Teaching — a Libraries research guide, last updated 1 September 2026; the AI detection section and the three model syllabus statements.
• Rutgers: Canvas external apps — Turnitin — the LMS integration.

This page is orientation, not legal advice. What binds you is Rutgers Policy 10.2.13 — in the official version at policies.rutgers.edu — the adjudication procedures for your campus, and your course syllabus.