Plagiarism and AI at Sydney: unacknowledged AI is filed under contract cheating

The short answer

At Sydney, handing in AI-generated content without acknowledgement or permission is not filed under plagiarism. It is filed under contract cheating — the same category as paying someone to write your essay. That classification decision is the single most important thing on this page, because contract cheating is where Sydney’s heaviest penalties live.

The practical consequence is simple. Find out, in writing, whether AI is permitted for each assessment and how it must be acknowledged — and then keep the evidence of how you worked. Sydney’s AI detector is switched on, but it is used only when someone already suspects something, and Sydney says in writing that its score is not the only evidence relied upon.

Where AI sits: contract cheating

Sydney’s plain-language policy page for its Academic Integrity Policy lists the key examples of a breach. Read the fourth item carefully:

“Breaches can be accidental or intentional. Key examples include: Recycling: Resubmitting your own past work without permission. Plagiarism: Using others’ work (text, images, ideas) without proper acknowledgement. Collusion: Improperly collaborating on individual tasks or assisting another student to obtain an academic advantage by dishonest or unfair means. Contract cheating: Having someone else complete your assessment, including submitting AI-generated content without acknowledgement or permission. Exam cheating: Any dishonest act in an exam. Fabrication: of data, information, images, or sources. Inappropriate technology use: Using tools like AI in ways not permitted or without citation. Unauthorised sharing…”

Two of the seven categories reach AI, and they are not the same. An AI tool used in a way the unit did not permit, or used without citation, is “inappropriate technology use”. AI-generated content submitted without acknowledgement or permission is contract cheating. The second is the more serious label by a distance.

That matters because of how Sydney describes contract cheating elsewhere. Its breaches page states plainly:

“Contract cheating is also illegal under Australian law.”

And on consequences: “The University takes contract cheating, misconduct and impersonation very seriously and applies severe penalties under the University of Sydney (Student Discipline) Rule. Engaging in contract cheating can result in failing your unit of study, or even suspension or expulsion from the University.”

Sydney’s human definition of contract cheating is also broader than most students expect, and money is explicitly irrelevant:

“Contract cheating involves getting someone else to complete all or part of your work and then submitting the work as if you had completed it yourself… There doesn’t have to be a payment for something to be considered contract cheating.”

The examples reach paid tutoring and model answers:

“You can be accused of contract cheating if you: buy a completed assignment from a ‘tutoring’ or ghostwriting company; ask a partner, friend or family member to write part or all of an assignment for you; pay a private tutoring company to coach you on how to complete an assignment; submit ‘model’ assignment answers provided by a private tutor or tutoring company; get someone to sit an exam for you; sit an exam for someone else; buy, sell or swap completed assignments or assignment answers via ‘sharing’ websites or social media platforms.”

If you use a paid study service of any kind, those two bolded lines are worth re-reading before you use its output.

Sydney also publishes two warnings that are unusual and worth taking seriously. First, on blackmail: “some contract cheating companies blackmail students who receive assistance, threatening to tell the university that the student has cheated unless a sum of money is paid by the student. This can even happen long after the student has graduated, which could lead to a qualification being revoked.”

Second, on companies selling help with your case: “Be cautious of external companies claiming to assist with academic cases. They falsely advertise high success rates, charge high fees, and may encourage unethical practices.” Sydney names free alternatives — see what an investigation looks like below.

The full list of breaches

Sydney’s breaches page gives the operative AI examples separately, and this is the list to check your own habits against:

“Unauthorised or inappropriate use of generative artificial intelligence. Misusing generative AI can breach the Academic Integrity Policy. Examples include: using generative AI such as chatbots, paraphrasing tools or translation toolsin an assessment where their use has been prohibited; submitting AI-generated work without appropriate acknowledgment; inputting University teaching or course materials, content generated by another student, Intellectual Property from external partners, or any person’s personal or health information into AI tools.”

Note that translation tools are named. If you draft in your first language and translate, that is a tool use, and in a unit where tools are prohibited it is caught by this clause. If you are writing in a second language, that is an uncomfortable rule and it is better to know it now than to discover it later. Ask your unit coordinator what is permitted, in writing.

Sydney’s statement of responsibility is broad too: “As a student, you are responsible for acting with academic integrity. This means submitting your own original work and correctly citing all sources, including from technologies like AI.”

There is one carve-out that is rarer than it should be, and it is worth knowing when you are worried about a similarity score:

“It is not considered plagiarism if: the ideas or words are commonly used and there is no other way to express them; you have made the discovery yourself through experimentation or analysis; you have combined the work and ideas of others to reach your own conclusion and have acknowledged these sources in the body of your work.”

The AI detector is on — on suspicion

First, the similarity check, which is universal. Sydney states that all text-based work goes through it:

“All text-based or written work that is submitted for assessment will be checked for plagiarism by being passed through text-matching or similarity-detection software such as Turnitin. Other assessment items, such as presentations or programming code may also be checked this way.”

Sydney is careful about what a similarity score means, and this is the calmest paragraph on its site:

“The similarity score denotes how similar your work is to other work within Turnitin’s database. There will always be some degree of text-matching when using Turnitin, which can occur as a result of your legitimate use of direct quotations, technical terms and phrases, and bibliographic material. If your assignment is referenced appropriately, this is not considered to be plagiarism. And, markers will easily be able to discern minor referencing issues from more serious issues, such as significant plagiarism.”

A non-zero similarity score is normal. It is not an accusation. More on that in what a Turnitin score actually means.

The AI writing detector is a separate thing, and Sydney describes its deployment precisely. The full statement, from the page that states it was last updated 2 June 2026:

“Detecting generative AI use. Turnitin has a tool for detecting use of artificial intelligence in student work. If a marker or teacher suspects that part or all of your assessment has been generated using AI technology and its use was not permitted, or use was not acknowledged, the Turnitin AI detection tool may be used to evaluate the situation. It’s important to note that the AI detector score would not be the only evidence relied upon for an academic integrity case, but will be considered alongside other relevant evidence.”

Three things follow from that wording, and each is useful to you.

It is suspicion-triggered, not universal. Unlike the similarity check, the AI detector is described as something a marker reaches for when they already have a concern. Something else came first — the writing, the marks, an inconsistency.

The score is corroborating evidence, not the case. Sydney has written down that it “would not be the only evidence relied upon”. If you are ever facing an allegation, that sentence is Sydney’s own, on Sydney’s own domain, and you may quote it back.

So the other evidence is what decides it. Which means drafts, version history, notes and your ability to talk about your own argument are the things that actually matter. See how to prove you wrote it yourself.

The contact point named throughout those pages is the Office of Educational Integrity (OEI), at educational.integrity@sydney.edu.au.

The two-lane model, and the contradiction

This section exists because Sydney publishes two positions that do not obviously fit together, and you deserve to see both rather than whichever one suits us.

Sydney’s Educational Innovation team publishes a “living FAQ” on the two-lane approach to assessment — authors Adam Bridgeman and Danny Liu, published 2 July 2024 and updated 21 April 2026. It is built on an explicit premise:

“it is already not possible to restrict AI use in assessments which are not secured in face-to-face environments. It is also not possible to reliably or equitably detect that it has been used, either because a student has access to the latest technology, because they know how to change the raw output, or can afford to pay someone to do this for them. Any unenforceable restriction damages assessment validity…”

So one part of the University runs the detector as corroborating evidence in investigations; another part of the University builds its assessment doctrine on the position that AI use cannot be reliably or equitably detected. Both statements are current and both are Sydney’s own. We are not going to resolve that for you — we have no standing to, and no measurement that would let us. We are telling you it exists, because if you are ever in a conversation about a detector score, the second quotation is on a University of Sydney domain.

The two-lane framework itself is worth understanding. Sydney’s table:

• Secure (Lane 1) — “Assessment of learning”, operating “Mainly at program level”, “Secured, in person”. Generative AI “May or may not be allowed by examiner”. Examples: “In person interactive oral assessments; viva voces; contemporaneous in-class assessments and skill development; tests and exams.” Mapped to TEQSA “Principle 2 – forming trustworthy judgements of student learning”.
• Open (Lane 2)— “Assessment for and as learning”, “Mainly at unit level”, “‘Open’ / unsecured”. “As relevant, use of AI scaffolded & supported”. Examples include using “AI to provoke reflection, suggest structure, brainstorm ideas, summarise literature, make content, suggest counterarguments, improve clarity, provide formative feedback”. Mapped to TEQSA “Principle 1 – equip students to participate ethically and actively in a society pervaded with AI”.

Sydney says the assessment principles in its Coursework Policy “follow the TEQSA ‘Assessment reform for the age of artificial intelligence’ and the Group of Eight principles on the use of generative artificial intelligence”. Australia’s regulator, TEQSA, required providers to file an AI action plan across 2024–25; a document identified as Sydney’s own action plan was not found on its public pages, but the two-lane framework above is published and is expressly aligned to the TEQSA principles.

The part with the sharpest practical edge is that Sydney has abolished the middle setting most universities use:

“In the two-lane approach, there is no ‘amber’ traffic lightand AI use can only be restricted when we can do so securely and reliably. In Open (Lane 2) assessments, there is only a ‘green’ light and the option of which tool to use for which activity becomes more like a menu with the educator helping to guide but not control the choice.”

And, asked directly whether a traffic-light scale can be applied in lane 2: “No.”

There is also a candid line about the transitional setting: “The drop-down item ‘unsecured assessment – no AI’ being available for Semester 1, 2025 is only a temporary measure, noting that it is actually not possible to enforce this.”

Be careful how you read that. It describes what the education portfolio thinks about enforceability. It does not repeal your unit’s rule. A restriction can be unenforceable in the aggregate and still be the condition you agreed to for your assessment — and unacknowledged AI content is still classed under contract cheating. Sydney itself says the realignment will take time: “Although we are aiming to align to the new assessment categories and types from Semester 2 2025, it is expected that more profound changes to the assessments themselves and the curriculum will occur over a longer timeframe of 1-2 years.”

Sydney tells you not to use outside checkers

This section works against our commercial interest, and it ships anyway, because you should hear it from us rather than in a hearing.

“You should avoid using third-party similarity or AI detection sites, as these sites may misuse your work without your permission.”

That is Sydney’s instruction, on its investigation page, and it covers tools like ours. The concern it names — what an outside site does with your text — is a fair one, and it is the right question to ask of any checker, ours included.

What we can say about our own tool is what we say everywhere: it asks for no name and no account, our published error rates are measured on English and German corpora, and no detector — ours included — produces proof. But Sydney’s rule is Sydney’s rule. If you study there, follow your unit’s guidance and ask your unit coordinator for a Turnitin report before you reach for anything external.

The three-rung ladder

Sydney sorts outcomes into four positions, and the differences between them are large. Verbatim, from the investigation page:

“No breach — There is no issue with your work and the matter is dismissed.”

“Minor breach — These are small breaches of academic integrity requirements that can reasonably be attributed to inadvertence or a failure to fully understand referencing requirements or acceptable academic practice. For example, not including quotation marks around a direct quotation or minor breaches of examination conditions.”

The outcome at that rung is educational: “you will be sent a letter (via email) outlining the steps to resolve the matter. You will be required to complete an approved development course and may receive a small, remedial mark reduction to remove any advantage that could be gained through inappropriate academic practice.”

Where it is handled matters too: “Minor breaches (such as those relating to low-level plagiarism or recycling of previously assessed material) may be determined locally by your unit coordinator. If you have a prior record of an academic integrity breach, the case cannot be handled locally by your unit coordinator and will instead be directed to the Faculty’s Educational Integrity team to be assessed.”

“Major breach — More serious forms of academic integrity breaches include persistent or reckless disregard for accepted academic practice. For example, making no attempt to acknowledge source material or fabricating citations. Two or more minor breaches can also constitute a major breach.”

That last sentence is the reason a minor breach is worth taking seriously even when the mark penalty is small. Two of them add up.

“Misconduct— Misconduct includes contract cheating, repeated academic integrity breaches, or breaches deemed by the decision maker to warrant a larger penalty than for those available for ‘major’ breaches. Misconduct has the same meaning as is outlined in the University of Sydney (Student Discipline) Rule, and will be dealt with in the same manner.”

Since unacknowledged AI content is classed under contract cheating, and contract cheating is named in the misconduct rung, you can see how a matter travels.

Intent is not irrelevant, though. Sydney draws the distinction openly:

“Not all acts of plagiarism are intentional. In some situations, it may be caused by your failure to understand the requirements of referencing. If this occurs early in your candidature and you have no prior breaches on record, you might be found to have committed a ‘minor breach’ only, and you will be directed to complete a development activity. You might also have a small mark penalty applied.”

And the aggravating factors: “Plagiarism will be considered more serious where you have done it on purpose, repeatedly, or if the amount of copied or unacknowledged work dominates your own original work.”

Sydney frames its whole approach the same way: “When integrity breaches occur due to a lack of understanding, the University generally adopts an educational approach. However, for more serious breaches, such as contract cheating, the penalties are more severe…”

On sanctions specifically, the enumerated schedule sits in the Student Discipline Rule and the Academic Integrity Procedures 2022. Both are served only through the Policy Register’s JavaScript viewer, and we could not read them. We are not going to summarise a penalty schedule we have not seen. What Sydney states on its readable pages is above.

Acknowledging AI, and keeping the outputs

There is no single university-wide AI declaration form on the pages we could read. The duty is set locally, per unit of study:

“Unit of study coordinators will specify if, and which, artificial intelligence (AI) tools are permitted for each assessment and how their use must be acknowledged.”

Sydney’s general acknowledgement duty is broader than AI, and it is worth reading if you have had a friend read your draft:

“Behaving with academic integrity also means that you must acknowledge where other people, technologies or tools have contributed to your work in some way. This could include acknowledging words, ideas, images, computer code, or artistic or musical creations. It can also mean acknowledging any assistance provided by other people in relation to your work, including any feedback, proof-reading or editing.”

Where AI is permitted, Sydney sets three student obligations, and the second is the one people skip:

“If AI use is permitted, keep track throughout your assignment process of all the ways you have used generative AI, and keep copies of generative AI outputs as evidence of your research and writing process.” · “Ensure that you take ownership for your final work.” · “Ensure that you have acknowledged any use of software or tools, and that you reference all sources.”

Keeping the outputs is protective, not incriminating: it is the record that shows what the tool did and what you did. Our AI disclosure statement guide covers how to word an acknowledgement — but at Sydney, follow the form your unit coordinator sets, because that is the binding one.

One more thing to check yourself: fabricated citations are named as an example of a major breach. Verify that every reference exists and says what you claim it says — verifying AI citations.

What an investigation looks like

Sydney publishes the route a case takes, which is more than many universities do:

“If a tutor, marker or other staff member suspects you’ve breached the academic integrity policy, they will usually go to the relevant unit coordinator who’ll assess whether there’s evidence of a breach, and if so, determine what type of breach.” · “Some minor breaches can be handled by unit coordinators locally while more serious breaches are referred to the relevant faculty’s Educational Integrity Coordinator… The most serious instances of suspected cheating are handled under the misconduct process, which is governed by the University of Sydney (Student Discipline) Rule and handled by the Registrar.”

The procedural fairness steps are set out, and they are your rights:

“Before determining if the allegation has been substantiated, the Faculty Educational Integrity team will give you written notice that: informs you of the allegation; provides instructions on how you should respond and a timeframe to respond; requests you to attend a meeting or submit a written response.”

“Once you have responded… the Faculty Educational Integrity team will consider your response and the evidence. All available evidence will be considered. This may include the work you submitted, similarity-detection software reports from Turnitin, source publications, examination incident reports and prohibited materials confiscated during an examination.”

“If you are requested to attend a meeting, it will usually take about 15 to 20 minutes and will include at least you, the Educational Integrity Coordinator from your faculty, and a professional staff member from the faculty. You can bring a support person…”

Fifteen to twenty minutes is short. Prepare for it: bring your drafts, your notes, your reading, and be ready to explain your argument out loud in your own words. See what to do when you are accused of using AI.

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.

Sydney names free, independent advocacy — take it, rather than paying one of the companies Sydney warns about:

“Undergraduate:contact the Caseworker Help service provided by the Students’ Representative Council. Postgraduate: contact the Postgraduate Advocacy Service provided by the Sydney University Postgraduate Representative Association (SUPRA).” — described by the University as “free, confidential and independent of the University.”

Reporting can be anonymous: “As a student, you can contact the Office of Educational Integrity at educational.integrity@sydney.edu.au to report an incident. You can do this anonymously.”

One timing point that catches people out. The current instruments took effect on 20 February 2023, and older work is judged under the older rules: “The former policy and associated procedures, which were superseded in February 2023, are the Academic Honesty in Coursework Policy 2015 and the Academic Honesty Procedures 2016. These will apply for any breaches related to submissions prior to 20 February 2023, and as such you will be provided access to copies of them when relevant.”

Finally, the compulsory module, which has a consequence attached that surprises people: “The Academic Honesty Education Module (AHEM) is mandatory for all commencing coursework students. It takes about an hour to complete. You need to complete it by the census date of your first semester. Your academic record can be suppressed if it isn’t completed.” Sydney adds that it “is regularly reviewed with new information, especially about new technologies such as AI”, and that completing it means “you won’t have to retake it for up to five years, even if you change courses.”

Can you check your own work first?

At Sydney this is discretionary, not guaranteed. The wording is “may”:

“Your unit coordinator may give you access to the Turnitin report when you submit an assignment. You can use this report to help improve essential academic skills, like proper citation and paraphrasing.”

So ask. If your unit coordinator enables it, you get the University’s own report on your own work, which is the best-placed evidence there is — and it comes without the third-party concern Sydney raises about outside sites. Note also that the report covers similarity; nothing on the pages we read offers students a self-service AI-writing check.

What Sydney does not publish where we could read it

We would rather name the gaps than paper over them, and one of them is a limit on us rather than on Sydney.

• The clause-numbered policy text was unreadable, not absent. Sydney’s Policy Register serves the Academic Integrity Policy 2022, the Academic Integrity Procedures 2022 and the Student Discipline Rule through a JavaScript PDF viewer that we could not fetch. Those documents exist and they bind you. Everything quoted on this page comes from Sydney’s plain-language policy pages and its Office of Educational Integrity student pages instead.
• The enumerated penalty schedule is in those unreadable documents. We state the outcome tiers Sydney publishes and no more.
• The standard of proof is not stated on any Sydney page we could read. It would sit in the Student Discipline Rule. We do not know it, so we do not assert it.
• No case counts. Sydney publishes no annual breach numbers that we found.
• No single prescribed AI declaration. The form is set per unit of study.
• External review— Sydney’s Academic Appeals Rule exists in the register but its text was not retrievable, and no external review route was documented on the pages we read.

Before you submit

• Get the AI rule for that assessment in writing. It is set per unit of study, by your unit coordinator — including how any use must be acknowledged.
• Treat paraphrasing and translation tools as tools. Both are named in Sydney’s examples of misuse.
• Keep drafts, notes and any AI outputs. Sydney tells you to keep the outputs, and drafts are what an investigation actually weighs.
• Check every reference exists.Fabricated citations are Sydney’s own example of a major breach.
• Ask your unit coordinator for the Turnitin report rather than using an outside site — Sydney advises against third-party similarity and AI detection sites.
• Take the free advocacy, not the paid kind. SRC Caseworker Help for undergraduates, SUPRA for postgraduates.

If you want a read on your own writing generally, our AI checkis free up to 1,500 words and asks for no name or account — but note Sydney’s explicit advice about third-party sites above, and follow your unit’s rule first. Our published error rates are measured on English and German corpora, and no detector — ours included — produces proof. On what those numbers can and cannot tell you, see how accurate AI detectors really are and, if you write in a second language, AI detectors and non-native speakers.

Sources

• University of Sydney: Artificial intelligence — the AI detection statement and the student obligations where AI is permitted; page states “Last updated: 02 June 2026”. Office of Educational Integrity.
• University of Sydney: Types of academic integrity breaches — contract cheating, the generative-AI misuse examples, and the intent distinction; “Last updated: 02 June 2026”.
• University of Sydney: Investigation process — Turnitin on all text-based work, the three-rung ladder, procedural fairness, the third-party checker advice; “Last updated: 02 June 2026”.
• University of Sydney: Academic integrity — the acknowledgement duty, the AHEM module, and the 20 February 2023 transition.
• University of Sydney: Academic Integrity Policy (plain-language page) — the Academic Integrity Policy 2022 superseded the 2015/2016 instruments on 20 February 2023. The clause-numbered PDF in the Policy Register is served through a JavaScript viewer and was not retrievable.
• University of Sydney: Assessment Procedures — unit of study coordinators specify which AI tools are permitted and how use must be acknowledged.
• University of Sydney, Educational Innovation: Frequently asked questions about the two-lane approach — Adam Bridgeman and Danny Liu; published 02/07/2024, updated 21/04/2026; a self-described “living FAQ”.

This page is orientation, not legal advice. What binds you is your unit of study outline, the University of Sydney Academic Integrity Policy 2022 and its Procedures, and the University of Sydney (Student Discipline) Rule.