Plagiarism and AI at UNSW: the six categories, and what counts as a tool

The short answer

UNSW does not have one AI rule. It has six, and your course outline tells you which applies to each assessment. Getting that right is most of the job.

UNSW is also unusual in this series for having AI detection switched on — while telling markers, in the same document, not to rely on it.

The six categories

These are the standard wordings as they appear in a course outline. Verbatim:

• No Assistance — “This assessment is designed for you to complete without the use of any generative AI.”
• Simple Editing Assistance — “you are permitted to use standard editing and referencing functions in the software you use to complete your assessment.”
• Planning/Design Assistance — “You are permitted to use generative AI tools, software or services to generate initial ideas, structures, or outlines.”
• Assistance with Attribution — “This assessment requires you to write/create a first iteration of your submission yourself. You are then permitted to use generative AI tools, software or services to improve your submission.”
• Generative AI Software-based Assessments — “This assessment is designed for you to use generative AI as part of the assessed learning outcomes.”
• Not Applicable — “Generative AI is not considered to be of assistance to you in completing this assessment.”

UNSW groups them: no AI permitted (the first two), AI permitted before the final artefact (the third), AI permitted in completing it (the fourth and fifth).

The fourth is the interesting one, and the one to read carefully. Under Assistance with Attribution you must write the first version yourself and may then use a tool to improve it — with attribution. That is a genuinely workable middle position, and it is more permissive than most universities in this series. But it only applies where your course says it does.

The duty to tell you sits with the university: “Students must be provided with clear instructions stating whether they can use ChatGPT or other forms of GenAI for each assessment or learning activity and if so, for what purpose. This notification needs to be provided to students in writing and through multiple channels.” If you cannot find it, ask — and keep the answer.

Where use is permitted, the standard is: “they must be properly credited by the student, but the submission must be substantially the student’s own work.” Our AI disclosure statement guide covers how to word that; UNSW publishes no single prescribed form.

What UNSW counts as a tool

This deserves its own section because UNSW’s list is broader than most students expect — and it names our own product category, which we would rather tell you than let you discover in a hearing.

The Plagiarism Policy’s own opening definition already reaches AI: “Plagiarism is presenting another person’s work, ideas or generative artificial intelligence output as your own.” Then, under Misuse of Generative AI tools, it lists examples:

“• Text generators • Translation tools • Proofreading/ polishing tools • Image, video, and art generation tools • Text spinners”

— used “to complete coursework or research without authorisation and/ or without appropriate referencing”, which “may be considered plagiarism”.

So at UNSW, a proofreading or polishing tool is in the same list as a text generator. That includes tools like ours. Our check marks filler, hedging and tired transitions in your own sentences and suggests plainer wording — nothing is rewritten for you — but if your assessment is No Assistance, the category you are in is what matters, not how light the touch is. Under Simple Editing Assistance the standard editing functions of your software are permitted; under Assistance with Attribution improvement tools are permitted with credit. Check which one you are in.

Two related definitions worth knowing. Inappropriate paraphrasing is “changing a few words and phrases while mostly retaining the original structure and/ or progression of ideas of the original information without acknowledgement” — which is what a text spinner does by design. And self-plagiarism catches repeat students: “Where a student is repeating a course, they should seek permission from the course coordinator before re-submitting, in whole or part, the same piece of assessment.”

Detection is on — with conditions

UNSW runs Turnitin’s AI writing detection, and restricts it to that one tool. The reasoning it publishes is worth reading in full:

“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 because: Only Turnitin has been approved by UNSW Cyber Security as protecting student privacy. The accuracy of other detection tools is extremely low. Even OpenAI have warned that their own detection tool sits at less than 30% accuracy. That’s an enormous error rate. Many small companies and individuals have developed various detection tools. However, these tools’ sites frequently lack clarity regarding the cookies and information they collect, their storage methods, and their data and privacy policies.”

We will be straight: that instruction covers us too. If you study at UNSW, the university has told staff not to upload student work to any platform other than Turnitin, and the privacy reasoning behind it is sound. What we can say about our own tool is what we say everywhere: our published error rates are measured on English and German corpora, we do not add your text to any third-party comparison database, and no detector — ours included — produces proof.

UNSW then limits what its own authorised tool can do:

“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 Plagiarism Policy itself puts the tool second in the order of evidence: assessment uses “academic expertise and judgement”, then approved technology tools “as a guide”, then the evaluation rubric, then procedural fairness. More on what these scores support in what a Turnitin score actually means and how accurate AI detectors really are.

One practical warning from the same page: “Turnitin papers are not allowed to be deleted as its considered violation of academic integrity.” Once submitted, it stays.

What markers are told to look for

UNSW publishes its sense-check list, which is unusually transparent and tells you exactly what a marker is weighing:

“Very general statements or generic answers; Rigid or formulaically logical arguments and answer structure; No spelling mistakes, typos or other grammatical errors; Fabricated names, dates, events, places; Incorrect or fake referencing – references that do not match the text; Inconsistent writing style throughout the assignment; […] Written assignment writing style that does not match the student’s in-class verbal or written work.”

And immediately after, the caution:

“These signs are not definitive proof of cheating using GenAI. It is important to review further and use academic judgment before taking any action.”

UNSW also notes that “some forms of formulaic writing can lead to false positives”. If you write carefully and in a second language, several items on that list describe good practice rather than misconduct — see AI detectors and non-native speakers.

Two of the signs are entirely within your control, and both are worth acting on: verify every reference exists in a real database (verifying AI citations), and do not let your style lurch between assignments.

One clause to be aware of: “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.” Academic judgement about quality can reduce a mark without any finding against you.

You get a meeting first

This is a real procedural protection and it is written into the procedure:

“Where misuse of generative AI is suspected, the Course Authority (or their nominee) and/ or the SSIA, should require the student meet with them to allow an opportunity for the student to demonstrate their learning and knowledge of the work they submitted. Should the student be unable to demonstrate their learning, the matter proceeds to assessment of level of AI use.”

The guidance tells staff to run it non-accusatorially: to ask “for copies of drafts of their assignment; whether the student can explain the steps undertook to complete the assessment; whether the student can explain orally the work completed”.

Which tells you exactly how to prepare: keep drafts, be able to describe your process, and be able to talk about your argument out loud. See how to prove you wrote it yourself and what to do when you are accused.

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.

The five-level rubric

UNSW publishes a plagiarism evaluation rubric with five levels — Poor scholarship, Minor, Moderate, Significant, Serious — each carrying an AI-specific descriptor.

The bottom rung is deliberately educational. Poor scholarshipmeans “a lack of academic rigour as a result of inexperience or carelessness rather than a deliberate attempt to pass off others’ work as one’s own”, and its outcomes are things like referral to Academic Skills, referral to the responsible-AI resources, and completing an online integrity module. If your case is genuinely inexperience, this is the band to argue for.

Schools handle poor scholarship through to significant. Serious cases must be referred to the Conduct and Integrity Office, and “any action or outcome that is outside of what is available below for the respective levels MUST be approved by” that office — so a school cannot improvise a penalty.

The stated principles are “Fairness / Consistency / Timeliness / Education”.

Contract cheating, and editing

UNSW treats contract cheating as a type of plagiarism, and its definition reaches further than paying someone:

“Engaging another person (either paid or unpaid) to complete work that is then submitted as the work of the student […] This includes circumstances where a student […] submits work they may have edited which was substantially the work of another person, or where a student […] prepares a draft that is substantially modified by another (beyond minor editing).”

“Beyond minor editing” is the line. A friend tightening your grammar is minor editing; a friend restructuring your argument is not — and money is irrelevant either way.

UNSW also contemplates that AI itself may fall inside the Australian legislation on academic cheating services: “GenAI could possibly fall under ‘academic cheating services’ if it produces a substantial part of work for a student that they were required to complete as original work themselves. That last part is important. Whether or not the use of AI is a form of cheating depends almost entirely on the instructions provided to students.”

Which returns you, again, to the six categories.

The equity condition

One clause that deserves to be better known, because it is a right rather than a restriction:

“You must ensure equitable access.When ChatGPT or other forms of GenAI are accepted as part of an assessment, academics must ensure that they are easily accessible for everyone. There must be no physical, geographical or financial restrictions on students’ use of the tool.”

If an assessment effectively requires a paid AI subscription, that runs against UNSW’s own published condition. Worth raising rather than quietly paying.

For context on where UNSW sits in the Australian debate: its Director of AI Strategy in Education has published a signed argument against the two-lane secure/open model — “The two-lane approach is brilliant at focussing the mind. But, like many binary options, it over-simplifies to make a point” — while keeping “supervised ‘lane one’ assessments at key points in the degree”, and preferring “conversations with students about their submitted assessments” to what it calls “technological surveillance”.

Before you submit

• Find the AI category for that assessment. There are six, and they differ across your courses.
• Count polishing tools as tools. The policy names them.
• Keep drafts. The first thing you will be asked for is copies of them.
• Check every reference exists.Fake referencing is on the marker’s list of signs.
• Ask before reusing your own work if you are repeating a course.

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 — but note UNSW’s own instruction about uploading work to other platforms, and follow your course rule first.

Sources

• UNSW: Plagiarism Policy — effective 13 May 2026; includes the Student Plagiarism Management Procedure and the five-level rubric. No version number or review date published.
• UNSW: AI assessment guidance — the six categories, the detection statement and the sense-check list; page carries no date.
• UNSW: Turnitin — the AI writing detection deployment; page carries no date.

This page is orientation, not legal advice. What binds you is your course outline and the UNSW Plagiarism Policy.