False accusation of AI misuse

Overview

Being accused of using AI (such as ChatGPT) in your assessed work is treated as a form of academic misconduct, usually falling under the same rules as plagiarism, collusion or “contract cheating”. The important point to understand at the outset is that the university has to prove its case against you, normally on the balance of probabilities. It is not enough for a tutor to have a “gut feeling” or for an AI-detection tool to flag your work. The burden is on the institution to show, with evidence, that you more likely than not used AI in a way that breached its rules. A false accusation is defensible, and there are established internal and external routes for challenging it.

Why AI accusations are often weak

AI-detection software (such as Turnitin’s AI indicator and similar tools) is known to be unreliable and produces false positives. It cannot definitively prove that a human did not write a piece of work. It also disproportionately flags writing by non-native English speakers, neurodivergent students, and people who write in a plain or formulaic style. Turnitin itself and many universities have acknowledged these limitations, and a detection percentage on its own is generally not regarded as sufficient proof of misconduct.

The Office of the Independent Adjudicator for Higher Education (the OIA, the ombudsman for students in England and Wales) has made this clear in its casework. In one published case summary on AI and academic misconduct, the OIA found in the student’s favour because the provider had not shown what evidence actually led its panel to conclude that the student had used AI to the extent of committing misconduct. The OIA recommended that the university reconsider the allegation afresh and let the student submit a new statement. The lesson is that a properly reasoned, evidence-based decision is required, and a conclusion that rests only on a detection score or an assessor’s impression is vulnerable to challenge.

Your key rights in the process

You are entitled to a fair procedure. In practice this means:

You should be told clearly and specifically what you are accused of, in relation to which piece of work, and what evidence is relied on. Vague allegations are improper.

You should be given the actual evidence against you, not just a conclusion. If a detection tool is being relied on, you are entitled to know that and to challenge it.

You should have a proper opportunity to respond, in writing and usually at a hearing or meeting, before any decision is made.

You should be allowed to bring supporting evidence and, in most institutions, to be accompanied by a students’ union adviser or a friend.

The decision must be reasonable and must be based on evidence that actually supports the finding.

If these procedural protections are not respected, that itself becomes a ground of appeal and, ultimately, a basis for complaint to the OIA.

Gathering your evidence

Because the university has to prove use of AI, your task is often to show a credible innocent explanation and to demonstrate your own authorship. The strongest defensive material tends to be evidence of your writing process. Consider gathering:

1. Draft versions of the work, ideally showing progression over time. Word and Google Docs retain version history and edit timestamps, which can show organic development of the document.

2. Your research notes, reading lists, annotated PDFs, library loan records and browser history relevant to the assignment.

3. Any outlines, mind maps or handwritten planning.

4. Metadata and file creation/modification dates.

5. Records of supervision meetings, emails to tutors, or forum posts where you discussed the work.

6. Evidence of your normal writing style, such as previous assignments, so you can show consistency (or explain differences).

If the allegation rests on a Turnitin or similar AI score, ask specifically what percentage was returned, what tool was used, and what other evidence the university relies on. You can then point to the well-documented unreliability and false-positive rate of such tools.

Responding to the allegation

Prepare a clear, calm written statement that does the following:

1. States plainly that the work is your own and that you did not use AI in a prohibited way.

2. Explains your writing process and points to the supporting evidence above.

3. Directly addresses whatever the university has put to you. If they rely on an AI-detection score, explain that such tools are not reliable evidence of misconduct and cannot prove non-authorship.

4. If there is any legitimate explanation for what triggered suspicion (for example you used a grammar or spell-check tool, translation software, or a reference manager, all of which may be permitted), set that out honestly and check it against your university’s specific policy on acceptable use of AI tools, which varies between institutions and even between modules.

Do not exaggerate or misstate anything. If you did use an AI tool for something you believed was permitted, it is far better to be candid and argue about whether it was within the rules than to deny it and be caught out.

The internal procedure and appeal

Every university in England and Wales must have a published academic misconduct procedure, and you should read your own institution’s version carefully because the detail differs. Typically there is an initial investigation or meeting, then a decision or a panel hearing, and then an internal appeal.

Appeals are usually only allowed on specified grounds rather than a simple rehearing. Common grounds, reflected in typical university rules, are:

1. The procedure was not followed correctly.

2. Fresh evidence exists that was not reasonably available earlier.

3. The finding of misconduct was not reasonable given the circumstances (in other words the evidence did not justify it).

4. The penalty is disproportionate or has a disproportionate impact on you.

For a false-accusation case, the “unreasonable finding” ground is usually central: you argue that the evidence did not meet the required standard of proof. Watch the deadlines closely. Appeal windows are often short, commonly around ten working days from the date the outcome is issued, so diarise the deadline as soon as you receive any decision.

Get your students’ union involved

Almost every students’ union has a free, independent advice service experienced in academic misconduct cases. They know your institution’s regulations, they can attend meetings and hearings with you, and they can help draft your statement and appeal. Contact them as early as possible. This is one of the most useful practical steps you can take and it costs nothing.

Completion of Procedures and the OIA

Once you have exhausted the university’s internal procedures, the university must issue a Completion of Procedures Letter. That letter is your gateway to complaining to the Office of the Independent Adjudicator for Higher Education, the independent ombudsman for students at higher education providers in England and Wales. The OIA scheme is free to students.

The OIA does not simply substitute its own academic judgement, but it will examine whether the university acted fairly, followed its own procedures, and reached a decision that was reasonable and properly evidenced. As its published casework on AI shows, it will find in a student’s favour where the provider cannot demonstrate the evidence that supported a misconduct finding. You normally have twelve months from the Completion of Procedures Letter to bring an OIA complaint, so again note the deadline. If the OIA upholds your complaint it can recommend that the university overturn the finding, reconsider it, remove penalties, or provide a remedy.

Wider context and any court route

The relationship between you and your university is contractual, and in principle serious procedural unfairness could be challenged in court, and a public university’s decisions can in some circumstances be susceptible to judicial review. In practice, courts expect students to use the internal procedures and the OIA first, litigation is expensive and risky, and the OIA route is designed precisely for these disputes. Court action is very much a last resort rather than a sensible first step.

Practical order of steps

1. Read your university’s academic misconduct policy and its policy on permitted use of AI tools, noting all deadlines.

2. Contact your students’ union advice service immediately.

3. Gather your drafts, version history, notes and other evidence of authorship.

4. Ask the university exactly what evidence it relies on, including any detection tool and score.

5. Submit a clear, evidenced written response and attend any meeting or hearing, ideally with an adviser.

6. If the finding goes against you, appeal within the deadline on the appropriate ground, typically that the finding was unreasonable and not supported by evidence.

7. If you exhaust internal procedures without success, obtain the Completion of Procedures Letter and complain to the OIA within its time limit.

Key points to remember

The university must prove misconduct on the balance of probabilities; you do not have to prove your innocence, although evidence of your writing process is very persuasive. AI-detection scores are not reliable proof on their own and are frequently challenged successfully. Fair procedure and properly reasoned, evidence-based decisions are required, and failures on either front are strong grounds for appeal and for an OIA complaint. Act quickly because deadlines are short, and use your students’ union, which is free and experienced in exactly this kind of case.

If you tell me which university you attend, what precisely you have been accused of, what evidence they say they have, and what stage the process has reached, I can give you more specific guidance tailored to your institution’s rules.