How to disclose AI use in a research paper or assignment

Direct answer: Disclosure means telling your marker, in writing, that you used a generative AI tool, which tool it was, and what you used it for. Universities call this an acknowledgement or a declaration. It is a transparency statement, not an admission of misconduct, when AI use is permitted, the acknowledgement is exactly what keeps your work honest and compliant.

Evidence: UK and Australian sector bodies frame responsible, transparent AI use as compatible with academic integrity rather than opposed to it (Russell Group, 2023; Quality Assurance Agency for Higher Education [QAA], 2023). University guidance is consistent: when AI tools are permitted, students must acknowledge their use, usually in a declaration or citation (University of Melbourne, n.d.).

Example: A Vietnamese undergraduate MAAS coached at an Australian university worried that admitting she used an AI tool to check her English would look like misconduct. Her mentor reframed it: a clear acknowledgement is what an honest student does, and the real danger is using AI quietly and hoping no one asks.

Direct answer: Often you do not, if the tool only fixed spelling, grammar, and punctuation on words you wrote. But the moment AI generates sentences, rewrites your argument, summarises sources, or produces ideas, you cross into use that must be acknowledged. When you are unsure, the safe rule every university gives is to declare it.

Evidence: University guidance distinguishes incidental use, such as basic spelling and grammar checking, from substantive use that shapes content, and advises students to default to disclosing assistance when in doubt (University of Edinburgh, n.d.; University of Melbourne, n.d.). The threshold can vary by course, so the tool's purpose, not its name, decides whether to declare it.

Example: A MAAS-coached student used an AI editor only to correct article and verb-tense errors across her essay. Her mentor confirmed this was incidental language help, but they still logged the tool and kept a copy of the prompts, so the decision was defensible if her marker ever asked.

Direct answer: Read three things before you use any AI tool: your university's academic-integrity policy, your course or subject guide, and the specific assignment brief. Permissions differ, some tasks allow AI freely, some allow it only for limited purposes, and some ban it entirely. If the rule is unclear, ask your subject coordinator before you submit, not after.

Evidence: Sector guidance leaves the detail to individual providers and assessments, so universities set their own permitted, restricted, and prohibited uses (QAA, 2023; Tertiary Education Quality and Standards Agency [TEQSA], 2023). University student guidance therefore tells students to confirm that their subject coordinator has authorised AI use, and how, before relying on it for assessed work (University of Melbourne, n.d.).

Example: A Vietnamese student assumed AI help was allowed because a friend in another subject used it. His MAAS mentor had him read his own assignment brief, which restricted AI to brainstorming only. He adjusted his workflow and wrote an acknowledgement that matched what his course actually permitted.

Direct answer: Put the acknowledgement where your university tells you, usually a short "Acknowledgements" or "Declaration" section, often at the start or end of the work. Name the tool, its publisher, the link, what you used it for, and how you edited the result. One clear, specific sentence per use is stronger than a vague catch-all line.

Evidence: University guidance asks students to list the AI tool, publisher, URL, and a brief description of how it was used, and many recommend keeping the prompts and a note on your modifications (University of Melbourne, n.d.; University College London, n.d.). "A declaration is a personal statement that you write to describe for readers how you have used AI tools and technologies to prepare an assessment" (University of Melbourne, n.d.). A model acknowledgement reads: "I acknowledge the use of [tool, publisher, URL] to summarise my initial notes and to proofread my final draft."

Example: A MAAS mentor coached a Vietnamese student to replace "AI was used in this essay" with a precise line naming the tool, stating she used it to brainstorm headings and check grammar, and confirming she wrote and revised all content herself, turning a vague disclosure into one a marker could trust.

Where you are unsure whether your own use of AI crosses the line your institution draws, the Academic Integrity Check reviews the work and tells you plainly, before a marker does.

Direct answer: An acknowledgement is a personal statement that you used a tool to help produce the work; a citation credits AI-generated content you actually included, with an in-text reference and a bibliography entry. Use an acknowledgement for assistance like proofreading or brainstorming, and add a citation when you quote or directly use AI output.

Evidence: University library guidance treats acknowledgement and referencing as related but separate steps: declare assistance in an acknowledgement, and reference specific AI output using your required citation style when it appears in your work (University College London, n.d.). Both depend on your course permitting the use in the first place (QAA, 2023).

Example: A Vietnamese student had pasted an AI-generated definition into her literature review. Her MAAS mentor showed her that this needed a citation, not just an acknowledgement, and helped her decide it was better to rewrite the definition in her own words and cite the original scholarly source instead.

Direct answer: Undisclosed AI use, where disclosure was required, is treated as an academic-integrity breach and can lead to penalties from a failed task to a misconduct finding. Universities are redesigning assessment around AI rather than relying on detection alone, so the safe assumption is that undeclared use will eventually surface, not pass quietly.

Evidence: Australian regulators stress that detection tools alone are unreliable and that assessment is being reformed to assure learning authentically and ethically (Tertiary Education Quality and Standards Agency [TEQSA], 2023). UK sector guidance similarly ties acceptable AI use to maintaining academic standards and integrity (QAA, 2023; Russell Group, 2023), which is why undeclared substantive use is a serious matter.

Example: A MAAS mentor reviewing a draft noticed a paragraph in the unmistakable voice of unedited AI output, alongside a reference that did not exist. Rather than risk a misconduct allegation, the student rewrote the passage in her own words, checked every source, and added an honest acknowledgement of how she had used the tool.

Direct answer: Decide your AI approach before you write. Read your course's rules, separate light language help from substantive use, keep a short log of which tools you used and the prompts, and verify every AI-touched fact yourself. Building this habit early turns the final acknowledgement into a two-minute task instead of a worry.

Evidence: Because ESL students often lean on AI for language support, university guidance is especially clear that they should default to disclosing more than incidental use and keep records of how tools were used (University of Edinburgh, n.d.; University of Melbourne, n.d.). Transparency, not avoidance, is what keeps that support within the rules.

Example: A MAAS mentor coached a Vietnamese student through the Outline → Draft → Final model with transparency built in: the outline noted where language tools were allowed, the draft logged each tool and prompt, and the final stage produced a precise acknowledgement matched to her university's format. The student stayed the author throughout, with the mentor advising rather than producing the work.