Turnitin at Australian Universities: What the Rules Say Happens Next

In Australia, three public layers may shape what happens after an AI flag, but they do different jobs. The national regulator, TEQSA, required every registered higher education provider to submit a gen AI action plan. Your own university's academic integrity procedure governs the individual case: what the alleged offence is, who decides it, what evidence is considered and when you must reply. Since 1 February 2025, the National Student Ombudsman has also been able to review how a provider ran a process, though it cannot make a new decision or review the academic mark itself. A Turnitin report may be part of the evidence, but it does not decide the case on its own.

If you have just been contacted about an AI flag, the single most useful hour you can spend is reading your own institution's academic integrity procedure end to end. Not the student-facing summary page. The procedure document, in the policy library, with a version date on it. That document tells you how many days you have, who is in the room, whether you can bring someone, and what you can appeal later.

Every document quoted here is linked, and each one is dated in the source notes. Policies get revised. Open the link and check the version before you rely on anything on this page.

On 3 June 2024, TEQSA issued a request for information to every registered higher education provider in Australia. The request page states what was asked for:

"The RFI asked all registered higher education providers to submit a credible institutional action plan, oversighted by appropriate governance mechanisms, to address the risk gen AI poses to award integrity."

Responses were due through the provider portal by 3 July 2024. TEQSA's follow-up note, published 5 July 2024, describes what happens to providers that did not answer:

"During the review period, we will also follow-up with any providers who have not made a submission, or whose plans are insufficient or not achievable, to understand the barriers and encourage them to meet their obligations. If a provider continues to fail to submit an action plan, we will consider whether it is appropriate to use regulatory tools."

Two things follow from this for you as a student. First, your provider has been asked, in writing, by its regulator, to have a governed plan for gen AI and assessment. That is a reasonable thing to ask about, and many universities publish the resulting guidance openly.

Second, and less comfortable: TEQSA regulates providers. It is not the place an individual student takes a complaint about being flagged. The National Student Ombudsman described further down is one external complaint route this article verified, with limits of its own.

Generalities about "Australian universities" are not much use when you are the one with an email in your inbox, so here is a real policy, quoted. Adelaide University's Academic Integrity Policy lists inappropriate use of AI as a named category of academic misconduct, defined as conduct "to subvert the aims of assignment or give the student an unfair academic advantage", and gives four examples:

"submitting work produced (or produced in part) by generative artificial intelligence as the student's own work · using information generated by artificial intelligence without acknowledgement or attribution · using artificial intelligence or digital tools to submit work which significantly misrepresents the student's level of competence · using digital tools to disguise plagiarism, collusion, copying, contract cheating or any other Academic Misconduct."

Read the third and fourth bullets slowly. Neither one is about which tool you opened. Both are about what the finished submission misrepresents. Under Adelaide's wording, "but I only used it for grammar" is not automatically an answer. It is also why the assessment brief for your specific course matters alongside the university-wide policy.

The standard of proof is defined in the accompanying Academic Misconduct Procedure:

"Balance of probabilities means the standard of proof to determine that academic misconduct has occurred. The decision maker must be satisfied that based on the evidence received, the academic misconduct was more likely than not to have occurred."

That procedure also carries an outcome table, and where AI sits in it is worth knowing. Inappropriate use of AI shares a row with plagiarism:

In Adelaide's table, the standard outcome for plagiarism or inappropriate AI use is a mark reduction up to zero for the assignment. The standard outcome for collusion is zero for the assignment, while contract cheating starts at zero for the course. Confirmed findings are recorded on an internal Academic Integrity Register, and the procedure states that outcomes "are not recorded on the student's academic transcript."

The same Adelaide procedure runs on five-working-day intervals almost everywhere. Five working days for the Academic Integrity Officer to examine the initial report. Five working days for you to say whether you will attend a meeting. Five working days for the outcome email after the meeting. If you attend, the meeting is usually held within ten working days of the notification.

Two provisions are worth knowing before anything is signed.

You may bring someone. "If a meeting is held, the student may bring a support person." If the case escalates to a Formal Inquiry, the panel includes "one representative from the Adelaide University Student Association" alongside the Dean Students and two academic staff.

An early resolution offer ends your appeal rights. The procedure offers "Early Resolution" in cases where an educational approach plus a warning or reduced outcome seems appropriate, and it is only available to students with no prior record. Then:

"When a student accepts an Early Resolution, a finding of academic misconduct will be recorded on the Academic Integrity Register. No further investigation will occur, and the student has no right to appeal the outcome."

Accepting can still be the right call. It is quick, and the outcome sits at the mitigated end. But it is a decision about your appeal rights taken under a five-working-day clock, and the procedure sets out four things the Early Resolution email tells you: that you may make an early admission, that you have five working days to accept, that you must complete the directed integrity education, and that the finding goes on the register. The loss of appeal rights is not one of the four. It is in the next paragraph of the procedure. Decline it and you go to a misconduct meeting instead, where the full range of outcomes in the table above is back on the board.

UNSW publishes its staff guidance on AI in assessment openly, and it is unusually direct about the tool's status. From the guidance page:

"To be clear: if Turnitin identifies the potential for AI writing in a response, this is merely a flag for an academic investigation. Markers should rely on their own professional judgement, not the AI detection tool."

The same page tells staff which tools they may use at all:

"UNSW only authorises the use of Turnitin's AI Writing Detection Tool for detecting improper AI use. Students' work should not be uploaded to any other platform."

The stated reasons are privacy approval and accuracy, and the page notes that many detector sites "lack clarity regarding the cookies and information they collect, their storage methods, and their data and privacy policies." If you are ever told a free online detector produced the evidence against you at an institution with a rule like this, that is a fact about the process, not about your writing.

One line on that page cuts the other way and you should know it exists:

"In the first instance, evidence of AI-generated work could justify a significant reduction in the mark for the assessment without the need to prove misconduct."

A mark can move without any misconduct finding, which means without the procedure, the meeting or the appeal that a misconduct finding would bring with it. If your mark dropped and nobody opened a case, that is the route you are on, and the document to read is your assessment and marking policy rather than the misconduct one. Our page on universities dropping AI detection covers the score side of this in more detail.

UTS puts its misconduct definitions in the Student Rules, which are public and numbered. Rule 16.2.1 lists examples of cheating, and two of them name detection software directly:

"Manipulate an assessment to avoid the UTS detection software." · "Use of a third-party service to bypass the UTS detection software."

The same list also covers "Copying or reworking any material (e.g. text, images, music, video) from generative AI tools, and claiming this work as your own without declaring use of the relevant tool" and "Using generative AI tools for assessment tasks where it has been specified that they are not to be used."

Those bypass clauses matter regardless of what you think of detectors. At an institution with wording like that, the question a panel can ask is no longer only "did AI write this", it is also "what did you do about the check". That is a separate allegation with its own evidence.

The appeal side of the UTS rules repays reading before you need it. Rule 16.13.2 gives you 20 working days from notice of the decision to lodge a written appeal. Rule 16.13.3 lists the only grounds:

1. the decision was based on a material misunderstanding of the Rules
2. the decision was based on a material mistake as to the facts
3. a failure of procedural fairness occurred, including failure to follow specified procedural requirements likely to have affected the outcome
4. fresh relevant evidence has become available that was not available or known to you at the time and would be likely to have affected the outcome
5. the penalty was manifestly excessive or inappropriate

Rule 16.13.4 allows the penalty to be stayed until the appeal time expires or the appeal is determined. The committee is three people, you may object to a member with cause, and if you produce genuinely new evidence the committee normally refers the matter back rather than deciding it itself.

The National Student Ombudsman commenced operations on 1 February 2025, after legislation passed on 28 November 2024. Its about page lists "disciplinary processes" among the issues it handles, and then draws the line that matters:

"The National Student Ombudsman cannot make a new decision for your matter. What we will do is consider the way a decision was made and can make recommendations to your higher education provider on how the decision or process could be improved."

The complaints page is more specific about both sides. In scope: "Providers complaint and appeal procedures", "Providers failure to clearly explain their decisions", and "Student enrolment and exclusion". Out of scope: "Complaints related to academic judgement, for example the grade you are awarded", and anything involving a provider that is not TEQSA registered.

So this is not a second appeal against the finding. It is a route for arguing that the process was unreasonable, unexplained or wrongly run. In practice that maps closely onto the appeal grounds above, which is why keeping every email and every dated document is worth the effort even when the internal process feels finished.

Four things about that service are worth stating flatly, all from the NSO's own FAQ:

1. Find the procedure, not only the summary. Search your university's policy library for "academic misconduct procedure" and note the version date. Check whether a student-facing page is a summary and whether the full procedure contains additional deadlines or rights.
2. Write down every deadline in it. Check whether each one uses working days or calendar days. Adelaide's procedure uses five-working-day intervals at several important stages, but your institution may use different periods.
3. Check what your assessment brief said about AI. The university-wide policy is the frame, the brief is the rule you are actually accused of breaking. Save a copy of it now, since course sites get archived.
4. Ask who may come with you. Check the procedure for a support person, student association adviser or advocate, and note any notice or permission requirement. Asking what the rule allows is not an admission of anything.
5. Collect your drafts, notes and version history before the meeting, not after. Our guide to the academic integrity meeting after an AI accusation covers what tends to carry weight.
6. Before accepting any early or informal resolution, find the sentence about appeal rights. If it says accepting ends them, decide with that in front of you.