← Back to blog

Plagiarism and AI at RMIT: checking your own work before you submit is written into policy

|10 min read

Check your writing for AI text — free

The first 1,500 words are free, with no sign-up. Every verdict shows how often it is wrong about verified human writing — a figure no other detector publishes.

We are building a writing workspace: your Word or LaTeX document, your PDFs beside it, and an assistant that can only cite what is actually in them — see it and get notified.

The short answer

At RMIT, running your work through matching software before you submit is not a nice-to-have. It is listed in policy as one of the ways students meet their share of the responsibility for academic integrity. Very few universities put it that way, and it is the most useful sentence on this page.

The second thing to know is a gap rather than a rule. RMIT names no detection vendor in its policy and publishes no position at all on AI writing detection. What it publishes instead is what it will ask you for: drafts, and an oral defence of your own work.

Self-checking is your responsibility, in policy

The Academic Integrity Policy sets out what students are expected to do, and the last limb is the one to act on:

“(9) Students share responsibility for maintaining academic integrity at RMIT; this is achieved through: … using assessment declarations, and content matching or authenticity software prior to submission …”

And the matching duty on the institution, so it is not a responsibility without a mechanism:

“(13) … ensuring that students and staff have access to content matching software or authenticity reports

Two things follow. First, if you cannot find the pre-submission check for your course, that is a fair thing to raise with your course coordinator — clause 13 puts the access obligation on the college, not on you. Second, if a matter ever arises, the fact that you checked before submitting is evidence of the care that clause 9 asks for.

We could not establish the mechanics — which tool students get, how many attempts, whether the report is visible to you. RMIT’s Learning Lab had a page titled “Understanding Turnitin”, and it now returns “Archived page — This content has been updated or relocated” and a 404. We mention that because the removal of a student-facing explainer is itself a small finding, not because it means the entitlement has gone: the entitlement is squarely in policy. Ask your course coordinator for the operational answer.

No vendor named, no position on AI detection

RMIT’s policy language is technology-neutral throughout. Its detection strategy, verbatim:

“(7) RMIT undertakes to detect and deter breaches of academic integrity through strategies including but not limited to: · manual detection by academic and teaching staff · using content matching or authenticity software · monitoring websites that enable academic integrity breaches to take place · the management and blocking of websites that breach academic integrity such as assessment file sharing, or illegal websites that offer contract cheating services.”

No product is named. Turnitin appears only in student-facing material — for instance in RMIT’s published “Real Stories of Plagiarism” transcript, read by student volunteers, where one recounts: “The lecturer also mentioned that other parts of my assignment were directly copied from a journal article. This was picked up by Turnitin. I was surprised. … I remember a student saying that if they stayed under a particular percentage of the Turnitin score, they wouldn’t- it wouldn’t count as plagiarism. But unfortunately, it’s still cheating.” (The stumble is in the original transcript.)

That student’s misconception is worth naming: there is no safe percentage. A similarity number is not a threshold you may sit under. If that idea is floating around your cohort, see what a Turnitin score actually means.

On AI detection specifically, RMIT publishes nothing.We read the Academic Integrity Policy, the Academic Integrity Procedure, the Student Conduct Policy, the Student Conduct Regulations and Schedule 1, as well as RMIT’s student academic-integrity page, its Learning Lab integrity module and its Library AI guide. None of them states whether an AI writing indicator is enabled or disabled at RMIT, and none gives a reason either way. The terms “AI writing detection” and “AI indicator” do not appear.

That is a documented absence rather than an inference from a page we could not open — but it is also not a promise. RMIT has not said it does not use such a tool. It has said nothing. If the answer matters to you, ask your course coordinator and keep the reply.

What RMIT looks at instead: drafts and an oral defence

Where the policy is quiet about tools, the Procedure is specific about method — and human judgement comes first:

“(4) Staff apply their professional and/or academic judgment in identifying potential breaches of academic integrity and making an initial assessment of the level of violation.”

The signals staff are told to look for are behavioural and comparative, and there is no detector anywhere in the list:

“(13) Contract cheating or ghostwriting may be detected in assessment tasks by identifying anomalies or inconsistencies in: the level of engagement with a course and the quality of assessment tasks; assessment content, language level, coherence, punctuation, grammar, or vocabulary between assessment tasks; reference or lack of reference to course materials or requirements, or assessment criteria; file naming conventions and metadata; formatting within a document and changes in layout between assessment tasks, or the use of referencing software or automatic formatting between assessment tasks.”

Two of those are worth pausing on. “Language level … between assessment tasks” means the comparison is against your own earlier work, not against a general model of how students write — which is a fairer comparison than a classifier, and a reason not to let your style lurch between pieces. And “file naming conventions and metadata” means the file itself is evidence: author fields, creation dates and edit history all travel with a document.

Then the two things staff may ask you for:

“(14) In determining whether a student has engaged in contract cheating, a staff member may: meet with the student and request the student to provide: evidence of the research and writing processes (for example, drafts) before final submission, and/or an oral defence of the work to demonstrate authorship; seek advice from a Senior Officer or Student Conduct.”

RMIT turns that into a standing instruction to students:

Keep drafts of your work. To help prove the authenticity and originality of your work, you should keep all draft versions of your work to show how your assessments were developed. These can be requested at any time during your program.

“At any time during your program” is the phrase to take seriously. Keep drafts for finished units, not only the current one. Practically, that means letting version history accumulate in whatever you write in, and keeping your notes and reading. Our guide on proving you wrote it yourself covers what a useful record looks like; and if a conversation is already happening, 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.

An oral defence is exactly what it sounds like: being able to explain your argument, your sources and your choices out loud, in your own words. If you have genuinely done the work, this is the part that goes well — provided you have reread your own thesis recently.

Where AI actually sits in the rules

RMIT is the most technology-neutral institution in this Australian set. The words “artificial intelligence”, “generative” and “AI” do not appear in the Academic Integrity Policy, the Academic Integrity Procedure, the Student Conduct Policy, the Student Conduct Regulations or Schedule 1. That does not mean AI is unregulated at RMIT. It means it is caught by wording that was written to catch anything.

Limb one — inside contract cheating. Schedule 1 to the Student Conduct Policy lists, as a form of contract cheating or ghost writing:

“vi. submitting an assessment task generated or altered by software tools or an unauthorised or unacknowledged algorithm

Read “generated or altered”. This limb is not limited to text a machine wrote from scratch; text your own hands wrote that software then altered is inside the words. And note where the limb sits: in contract cheating, not in plagiarism. At RMIT, that is the more serious neighbourhood.

Limb two — the breach list. The Academic Integrity Policy includes among its examples “‘washing’, or the use of software services to disguise plagiarism” and “misuse or unauthorised use of technology or equipment”.

The plain-English version is on RMIT’s student webpage, not in the instruments — a distinction worth keeping in view, because the policy is what binds and the webpage is what explains:

Intentional plagiarism– when you deliberately use someone else’s ideas, words or images and present them as your own, including failure to appropriately and accurately acknowledge the use of Artificial Intelligence tools.” · “Using an AI tool when it’s not permitted could be considered ghostwriting and a breach of academic integrity.” · “Using AI in these ways is a breach of Academic Integrity Policy and may have serious consequences.”

RMIT’s misconduct definition, meanwhile, reaches further than most students expect:

Academic misconduct — Conduct by a student that is intended, or likely to have the effect of obtaining, for that student or any other person, an advantage in the performance of assessment by unauthorised, dishonest, unethical or unfair means whether or not the advantage was obtained.”

The attempt is enough, and the advantage need never materialise. Note also the alternative in the opening words: conduct that is “intended or likely to have the effect of” gaining an advantage. Intention is not required for the definition to bite — though, as the next section shows, it matters a great deal to which process you end up in.

RMIT’s general stance on AI is not hostile, for what it is worth: “AI is rapidly transforming the way we live, work and learn. We want to prepare our students for careers where AI capability is increasingly expected. We’re doing this by helping students develop both the practical skills, as well as the judgment to apply it ethically and responsibly.” RMIT also provides its own tool: “Val is an RMIT supported, secure and free-to-use generative artificial intelligence tool that functions just like ChatGPT.” Whether you may use it in a given assessment is still a course-level question.

Australia’s regulator, TEQSA, required providers to file an AI action plan across 2024–25. A document identified as RMIT’s was not found on its public pages, and RMIT publishes no secured/open “two-lane” assessment taxonomy. Its instruments do treat supervised assessment as its own category with its own limb — “Attempting to Gain Unfair Advantage in a Supervised Assessment — (22) Students breach academic integrity when they use any method to attempt to gain an unfair advantage in an invigilated or supervised assessment or test…” — but that is not a lane framework and we do not describe it as one. RMIT does link students directly to TEQSA’s own materials on protecting academic integrity and on illegal cheating services.

“Washing”, and tools that change your words

This section is against our interest, and it ships anyway, because you should hear it plainly rather than discover it.

RMIT is the only institution in this Australian set that gives paraphrasing and spinning services their own term in policy. From the breach list:

‘washing’, or the use of software services to disguise plagiarism

Schedule 1 repeats it as “using software services to disguise plagiarism (washing)”. Put that beside limb (vi) — an assessment task “generated or altered by software tools” — and the shape of the rule is clear: software that rewords your text so that it reads differently is the thing being named.

So, straightforwardly: a paraphrasing tool, a spinner, a rewriter or a “humanizer” is what these two clauses are about. If you are considering one to bring a similarity score down, that is precisely the conduct the word “washing” describes.

Our own product is a check, not a rewriter — it marks filler, hedging and tired transitions in your own sentences and suggests plainer wording, and nothing is rewritten for you. We would still rather you knew that RMIT writes about “software tools” altering an assessment task in the contract cheating limb, and that our category sits close enough to that line to be worth thinking about. Our published error rates are measured on English and German corpora, and no detector — ours included — produces proof.

One more thing in the same family, from RMIT’s July 2025 amendment record: clause 16 of the Procedure was amended “to clarify that the Course Coordinator provides authorisation for the use of professional editor”. At RMIT, even a human editor needs authorisation. If a paid human editor needs sign-off, do not assume software does not.

Contract cheating and Commonwealth law

RMIT ties contract cheating to Australian federal law more insistently than anyone else in this set. The phrase “not permitted under Commonwealth law” appears five times in a single limb of Schedule 1:

“b. contract cheating or ghost writing, including: i. the use or attempted use of course materials or assessment tasks from sources or providers whose services are not permitted under Commonwealth law, whether prewritten or specially prepared for the student concerned; ii. obtaining course materials or assessment tasks from sources or providers whose services are not permitted under Commonwealth law; iii. providing course materials or assessment tasks to sources or providers whose services are not permitted under Commonwealth law; iv. submitting an assessment task produced by another person, including a friend, family member, fellow student or a staff member of the University; v. undertaking an assessment task for another person; vi. submitting an assessment task generated or altered by software tools or an unauthorised or unacknowledged algorithm; vii. providing a student’s own assessment task, or that of another student, to sources or providers whose services are not permitted under Commonwealth law, or viii. encouraging or persuading another person to advertise the services of, or use or provide course materials to, a source or provider whose services are not permitted under Commonwealth law regardless of whether the use or sharing eventuated.”

Contract cheating is separately regulated under the Commonwealth TEQSA legislation in Australia, which is the law those five references point at. Three limbs here catch conduct students rarely think of as cheating at all:giving your own work to a study-sharing site (vii), uploading course materials to one (iii), and telling a classmate about a service (viii) — the last of which bites “regardless of whether the use or sharing eventuated”.

RMIT’s definition also explains, candidly, why matching software is not the answer to this:

Contract cheatingis the use of outsourced material for the purpose of submission by a student for assessment. … Students may obtain assessments from peers or ‘tutors’ and the arrangement may not involve a financial exchange. Significantly, the submitted work is usually original in nature making it difficult to detect using text matching plagiarism software.

That sentence is the reason RMIT’s method is drafts and an oral defence. Original text defeats a matching tool by construction, so the University asks about provenance instead.

And a warning worth repeating to anyone tempted: “Contract cheating services have been known to blackmail students in exchange for more money and personal information, such as their student login details. What starts as a short-term ‘solution’ can quickly lead to extortion, with serious academic, financial and personal consequences.” Reports go to student.conduct@rmit.edu.au.

The acknowledgement template — including the negative one

RMIT Library publishes acknowledgement and referencing guidelines for AI tools (the guide states “Last Updated: Jul 29, 2026”, and notes that “These guidelines were last updated in February 2026 in response to updated APA 7th referencing guidelines for generative AI”). The duty first:

“If you use any AI tools in your assessment, you must acknowledge the use of these tools and reference their outputs. … Failure to acknowledge and reference the use of AI tools can result in academic misconduct.

Importantly, the duty covers tools used in the process, even where no AI text reaches your final document:

“In some assessment tasks, you may be permitted to use AI tools for certain purposes, such as to: conduct background research on a topic; generate or brainstorm an outline or ideas for your assessment; clarify your reading interpretation or confirm your understanding of a topic; receive feedback for your writing (e.g. checking for consistency, clarity, grammar, spelling and expression).” · “In such cases, you will not include AI-generated content directly in your final work, e.g. you are using AI tools in the process of creating the work. For this type of AI use, however, you need to include an acknowledgement of what AI tools were used and how they were used within the body or methods section of your work.

Read that second quotation carefully if you use any writing-feedback tool, ours included: “receive feedback for your writing” is named as a use that requires acknowledgement where it is permitted. Using a checker and saying nothing is not a neutral act at RMIT.

The prescribed form:

“In your acknowledgement, include the following: how you used the AI tool; the name of the AI tool and its creator; the year the AI content was generated by the user. In your reference list or bibliography, include an entry for the AI tool.” · “Format for the acknowledgement (based on APA 7th, you may need to modify this for your referencing style): I used [tool used] (Creator, year) to [explain how tool was used].

RMIT’s own examples: “I used ChatGPT (OpenAI, 2026) to provide topic suggestions for my research project. I used the Writing Help persona in Val (RMIT, 2026) to provide feedback on my assignment draft.”

There is also a negative declaration, which is unusual and useful:

“If your course permitted you to use generative AI tools in your assessment, but you have chosen NOT to use them, it is suggested you put the following statement in your general acknowledgement: ‘I did not use any AI tools in the creation of this work.’

If you wrote it all yourself, say so. It costs one line and it puts your position on the record at the moment of submission rather than months later.

And keep the receipts:

“Keep a record of how you have used AI tools in the process of creating your work, including the prompt used, the outputs generated, and how the generated content was used in your work. Make sure you save a copy of the prompts used and outputs generated, as you may need to provide this to your educator on request.

For how to word an acknowledgement inside a thesis specifically, see our AI disclosure statement guide — but use RMIT’s template where RMIT gives you one. And verify your sources independently, because fabricated references are the easiest thing to be caught by: verifying AI citations.

Breach or misconduct: how RMIT splits them

RMIT’s split is explicit and it sits in the Procedure, which makes it easier to argue about than most:

“(5) First-time, low-level breaches of academic integrity characterised by inexperience, lack of student knowledge, or poor writing skills may be managed as an educational opportunity initially by a course coordinator.

“(6) Repeat or higher order breaches that may be intentional, deliberately planned, organised or systematic, are reported to a Senior Officer to manage as a potential instance of academic misconduct under the Student Conduct Policy…”

“Poor writing skills” being named in clause 5 is worth noting if you are writing in a second language: RMIT contemplates that weak writing, not dishonesty, is sometimes the explanation.

The factors that decide which side you land on are published:

“(9) … the level of knowledge and experience the student has by reference to their university journey; the extent of academic integrity training, including referencing practices relevant to the course; the student’s intention; any prior instances of academic integrity breaches and the educative response taken, and any other relevant circumstances giving rise to the failure.”

If your case is genuinely inexperience, those are the points to make, with evidence: where you are in your degree, what training you had, and what you were actually trying to do.

For completeness, RMIT’s definition of academic integrity itself: “(5) Academic integrity means that staff and students at RMIT act with the core values of honesty, trust, respect, fairness, and responsibility in education, learning, teaching, training and research, including by acknowledging the sources of ideas; both original and the work of others.”

The published penalty scale

At course-coordinator level — a breach, not misconduct — the Procedure caps what can happen to your mark:

“(10) Where a Course Coordinator (or equivalent) finds that a lack of understanding is the basisfor the student’s failure to appropriately and accurately acknowledge their sources, they must treat the instance as an assessment matter and: · explain that it is a breach of academic integrity · require the student to repeat the assessment task and/or deduct marks from the assessment (up to a maximum of 20% of the mark for the assessment task) · [for Vocational Education tasks with no numerical grading, deduction is not available and the task must be repeated] · direct the student to engage in educational support services · provide a written warning to the student advising that another breach of academic integrity will be handled as an allegation of academic misconduct and may result in more severe penalties, and · advise the relevant school/industry cluster academic services team to record the warning.”

A capped 20% deduction and a written warning is the low rung. The warning is the part that matters, because it changes the route next time.

At Senior Officer level, the Student Conduct Regulations list: a reprimand; suspension of enrolment “for a period not exceeding one compulsory semester or nominated teaching period”; repeating all or part of an assessment; recording “a failure or capped result for a course”; recording a failure for part of a task; paying for damage to RMIT property “up to a maximum of $AUD200 in total”; a verbal or written warning; a required educative or mediation activity; a combination; or other appropriate actions.

There is a genuinely useful boundary clause on what that officer cannot do:

“(44) A Senior Officer may not: cancel any or all final results; impose a financial penalty on the student…; revoke the student’s licence to attend campus…; exclude the student from using any of RMIT’s computing and network facilities for a period…; refer to Council a recommendation to revoke an award; or expel the respondent student.”

The Student Conduct Boardhas the heavier powers: everything a Senior Officer can do, plus cancelling any or all final results, fines under RMIT’s schedule of fees and charges, an order that fees are forfeited, revoking campus access, excluding you from computing facilities, suspension “for a period not exceeding two compulsory semesters”, referring to Council a recommendation to revoke an award, and expulsion.

Note the structure there: degree revocation is not a University-level power at RMIT — it is a recommendation referred to Council.

RMIT’s own student-facing summary is blunter: “If you breach academic integrity (even if accidental), this means that you are breaching RMIT’s Academic Integrity policy. This can lead to serious consequences, including: the forfeiture (loss) of results; failure; suspension; expulsion from your program.”

Undischarged penalties have knock-on effects: clause 35 of the Student Conduct Policy provides that where a student has not complied with a direction, penalty or consequence by the prescribed date, RMIT will withhold services, subject to the discretion of the University Secretary and Academic Registrar.

Appeals reach the merits

The standard of proof is stated expressly in the Student Conduct Policy: matters are “determined utilising ‘the balance of probabilities’ as the requisite standard of proof to be applied when determining matters under this policy and corresponding procedures”. It is also a defined term in the policy.

The chain runs: Course Coordinator (breach level) → Senior Officer → Student Conduct Board → Student Conduct Appeals Committee. Senior Officers “are predominantly drawn from positions reporting to Executives or Senior Executives of RMIT, upon the advice of the University Secretary and Academic Registrar”.

The important feature is what an appeal may challenge. The Appeals Committee “may consider and make determinations in response to eligible student appeal submissions that have been lodged against decisions (findings of fact), or the consequences or penalties applied by: a Senior Officer, or [the Student Conduct Board]”. So you can appeal the finding itself, not only the penalty. Clause 26 confirms that students may request a hearing, subject to eligibility.

External review is expressly preserved:

“(46) Nothing in these Regulations is intended to preclude a student from exercising any right to external review of any determination. (47) RMIT may pause any internal process under these Regulations where the student exercises the right to external review of a determination involving the same or similar subject matter, or where the matter is otherwise subject to legal proceedings.”

RMIT does not name a particular external body in the documents we read; it preserves the right generically and directs complaints through its own complaints channel. We are not going to name a body RMIT has not named.

There is also an academic freedom carve-out worth knowing: “(9) The exercise of academic freedom and freedom of speech by a student, subject to the limitations detailed in the Intellectual Freedom Policy, will not constitute misconduct nor attract any penalty or other adverse action by reference only to its content.” And a support duty: “(24) Students will be advised of support services available to them as part of any academic integrity breach report, investigation or outcome…”

RMIT publishes no case counts. It maintains an internal pipeline — confidential breach data, school notifications to the Centre for Education Innovation and Quality, and “scheduled reporting on trends and strategies to mitigate breaches of academic integrity to governing bodies” — but no numbers. What it does publish is anonymised narrative case studies: “Real stories of plagiarism, collusion and contract cheating … read by former or current student volunteers”, with full transcripts.

Before you submit

  • Run the pre-submission check. Clause 9 lists it as part of your responsibility, and clause 13 puts access on your college.
  • Keep every draft — for finished units too. RMIT says they “can be requested at any time during your program”.
  • Acknowledge tools used in the process, not only text that made it into the document. RMIT names writing feedback as a use requiring acknowledgement, and gives you the template.
  • Use the negative declaration — “I did not use any AI tools in the creation of this work” — if AI was permitted and you did not use it.
  • Do not use rewriting or “washing” software.It is named in the breach list by that word, and altering an assessment task with software tools sits inside RMIT’s contract cheating limb.
  • Get authorisation for a professional editor. The Course Coordinator provides it.
  • Never upload your own or another student’s work to a study-sharing site. Providing work to those providers is its own limb.

If you want a read on your own text before you hand it in, our AI check is free up to 1,500 words and asks for no name or account. Three honest caveats: our published error rates are measured on English and German corpora; no detector — ours included — produces proof; and at RMIT, if you use any tool to get feedback on your writing in a task where tools are permitted, the Library guidance says to acknowledge it. If you are worried about how your writing reads to a detector generally, see how accurate AI detectors really are and AI detectors and non-native speakers.

Sources

  • RMIT University: Academic Integrity Policy — Current; effective 11 July 2025; review 2 August 2026; approved by Academic Board 21 March 2022; owner: Sherman Young, Deputy Vice-Chancellor Education. Source of clauses 5, 7, 9, 13, 15 and the definitions of academic misconduct and contract cheating.
  • RMIT University: Academic Integrity Procedure — Current; effective 11 July 2025; review 2 August 2026; approved by the Deputy Vice-Chancellor Education 4 August 2021. Source of clauses 4, 5, 6, 9, 10, 13, 14, 22 and 24. The 11 July 2025 republication record notes the amendment to clause 16 on authorisation for a professional editor.
  • RMIT University: Student Conduct Policy — Current; effective 18 August 2026; review 1 January 2027; approved by Academic Board 25 October 2021. Source of the balance-of-probabilities standard (cl 10), the decision-maker chain and cl 35.
  • RMIT University: Student Conduct Regulations — Current; effective 1 January 2022; review 31 December 2027; approved by RMIT University Council 16 December 2021. Source of the penalty scales (cll 43, 45), the limits on a Senior Officer (cl 44), the academic freedom carve-out and the external review clauses (cll 46–47).
  • RMIT University: Student Conduct Policy Schedule 1 – Student Misconduct (PDF, version 6) — the contract cheating limbs, including limb (vi) on software tools and algorithms, and “washing”.
  • RMIT University: Academic integrity (student page) — the plain-English AI statements, the keep-drafts instruction, the blackmail warning, Val, and the “Real Stories” transcripts. This is student-facing guidance, not a policy instrument.
  • RMIT University Library: Artificial intelligence (AI): acknowledgement and referencing guidelines — the acknowledgement template, the negative declaration and the record-keeping duty; the guide states “Last Updated: Jul 29, 2026” and that the guidelines were last substantively updated in February 2026.

This page is orientation, not legal advice. What binds you is your course guide, the RMIT Academic Integrity Policy and Procedure, and the Student Conduct Policy with its Schedule 1 and the Student Conduct Regulations.

Frequently Asked Questions

Can I check my own work before submitting at RMIT?

Yes — and RMIT frames it as something you are responsible for doing. Academic Integrity Policy clause 9: “Students share responsibility for maintaining academic integrity at RMIT; this is achieved through: … using assessment declarations, and content matching or authenticity software prior to submission.” Clause 13 places the matching obligation on colleges: “ensuring that students and staff have access to content matching software or authenticity reports.” The mechanics — which tool, how many attempts — are not published in a page we could read, so ask your course coordinator.

Does RMIT use AI detection?

RMIT publishes no position, either way. We read the Academic Integrity Policy, the Academic Integrity Procedure, the Student Conduct Policy, the Student Conduct Regulations and Schedule 1, plus RMIT's student academic-integrity page and its Library AI guide: none states whether an AI writing indicator is enabled or disabled, and none gives a reason either way. RMIT's published detection strategy is generic — “content matching or authenticity software” — plus staff professional judgement.

Is AI mentioned in RMIT's academic integrity policy?

Not by name. The words “artificial intelligence”, “generative” and “AI” do not appear in RMIT's policy instruments. AI is caught by two generic limbs instead: “submitting an assessment task generated or altered by software tools or an unauthorised or unacknowledged algorithm” (Schedule 1, inside contract cheating), and “‘washing’, or the use of software services to disguise plagiarism” (Academic Integrity Policy cl 15). RMIT's student webpage does say it plainly: “Using an AI tool when it's not permitted could be considered ghostwriting and a breach of academic integrity.”

How do I acknowledge AI use at RMIT?

RMIT Library publishes a template: “I used [tool used] (Creator, year) to [explain how tool was used].” It also asks you to include an entry for the tool in your reference list, and to keep “a copy of the prompts used and outputs generated, as you may need to provide this to your educator on request.” If your course permitted AI and you chose not to use it, RMIT even suggests a negative declaration: “I did not use any AI tools in the creation of this work.”

Check your writing for AI text — free

The first 1,500 words are free, with no sign-up. Every verdict shows how often it is wrong about verified human writing — a figure no other detector publishes.

We are building a writing workspace: your Word or LaTeX document, your PDFs beside it, and an assistant that can only cite what is actually in them — see it and get notified.