This article is part of The Complete Guide to Using AI for Australian University Study, our deep-dive hub covering policies, tools, citations and what’s actually allowed at Australian unis.

You’ve got an assignment due next week, and you’ve used Claude to explain a complex concept or ChatGPT to help brainstorm ideas. Now you’re staring at your laptop, wondering: do I need to cite this? And if so, how the hell do I do it properly?

You’re not alone. Across the Australian university AI policies I’ve checked (QUT, UQ, Sydney, UNSW, Melbourne, ANU, and Swinburne), most now ask you to acknowledge AI use through a declaration, cover sheet or appendix rather than a formal reference-list entry, though a few (like UNSW) still expect formal citation when you’re directly quoting AI output. Always check your own institution’s current guidance before you submit.

Here’s the practical guide to citing AI tools without stuffing it up.

The Two Types of AI Acknowledgement

Before we dive into formats, you need to understand there are actually two different ways universities want you to acknowledge AI use.

Citation goes in your reference list. You’re treating the AI tool like a source: a book, journal article, or website that contributed to your work.

Acknowledgement goes in your methodology section, appendix, or author note. You’re being transparent about your process, like acknowledging a research assistant or editor.

Across the seven unis named above, the common pattern is acknowledgement, not citation. UNSW and QUT are the mixed cases: both accept an acknowledgement statement for tool-assisted work, but both also expect a formal in-text citation when you’re directly quoting AI output. TEQSA’s student advice on AI makes clear that where AI use is permitted, you need to understand how to reference it, and the rules differ between disciplines. Check your unit guide or ask your teaching staff if you’re unsure.

Common Acknowledgement Requirements

When you’re acknowledging AI use (the more common requirement), universities typically want four things:

  1. What tool and version you used (Claude Sonnet 4.5, ChatGPT GPT-4, Copilot, and so on)
  2. What prompts you gave it (the actual questions or requests)
  3. What you used the output for (brainstorming, explanation, rubric decoding, editing suggestions)
  4. How you verified the information (cross-checking sources, applying your own analysis)

Here’s a template that works for most Australian universities:

“I used [AI tool name and version] on [date] to [purpose]. My prompts included ‘[example prompt]’ and ‘[example prompt]’. I used the outputs to [specific use], and verified all factual claims against [sources/methods].”

Here’s a made-up example to show the shape of an attribution note. Imagine a QUT IT assignment, a systems-design task where the rubric’s HD criterion says something like “demonstrates sophisticated justification of architectural decisions with explicit consideration of alternatives.” You read that and think: what does “sophisticated” actually mean in practice? How is it different from Credit-level?

You paste the brief and the full rubric into Claude and ask: “What would an HD response to this criterion demonstrate that a Credit response wouldn’t?” Claude comes back with something like: “An HD response would explicitly compare at least two viable alternative architectures, justify the chosen one with concrete trade-offs (performance, cost, maintainability), and reference the course materials that inform those trade-offs. A Credit response typically describes the chosen architecture without comparing it to alternatives.” Now you know what you’re aiming at. You write the assignment yourself, but with a clearer target before you write a single paragraph.

Your attribution note at the end of the submission might read:

AI acknowledgement: I used Claude Sonnet 4.5 (Anthropic) during the planning phase of this assignment to help interpret the rubric’s HD criterion for system design. Specifically, I pasted the assessment brief and the rubric and asked the model to clarify the distinction between HD-level and Credit-level responses. I used the model’s explanation to inform my own approach but wrote all the content myself. All technical claims in the submission were verified against the course materials for the relevant week.

That’s the basic shape. Every university has its own preferred attribution format, so follow your own institution’s guidance on wording and placement. But the principles are the same: name the specific model and version you used (not just “AI” or “Claude”), be specific about what the AI did, be specific about what you did, and verify any factual claim the AI contributed to against a primary source.

Why the specific model version matters: AI model behaviour drifts across versions. A response from Claude Sonnet 4.5 in March 2026 is not the same as a response from an earlier or later version, and reproducibility depends on naming the exact model. “I used Claude” is too vague. “I used Claude Sonnet 4.5” is the standard to aim for.

APA 7th Edition: The Full Citation Format

If your unit requires APA referencing and wants a full citation (not just acknowledgement), treat AI tools like software. APA’s official guidance on citing ChatGPT confirms this approach. The format is:

In-text citation: (Anthropic, 2026)

Reference list: Anthropic. (2026). Claude (Sonnet 4.5 version) [Large language model]. https://claude.ai

Replace the details as needed:

Include the version or date you accessed it. AI tools update constantly, so your March 2026 conversation might give different answers than someone else’s April 2026 conversation.

Harvard Referencing: The Alternative Format

Many Australian universities use Harvard referencing. The format is similar but slightly different:

In-text citation: (Anthropic 2026)

Reference list: Anthropic 2026, Claude (Sonnet 4.5 version), large language model, accessed [date], https://claude.ai

University-Specific Requirements

Here’s where it gets tricky. Many universities have their own specific requirements that override the standard referencing guides.

University of Queensland has one of the most comprehensive Australian guides to AI citation. Their library guide covers both acknowledgement and formal citation for AI use, with specific templates provided for APA, Harvard, and other referencing styles.

University of Sydney’s AI policy requires acknowledgement of AI use in open (unsupervised) assessments, including the tool name, version, publisher, URL, and a brief description of use. Supervised assessments generally prohibit AI unless stated otherwise.

The key point: check your university’s specific requirements. A quick email to your librarian or teaching staff can save you from getting the format wrong.

What Prompts to Include

You don’t need to include every single prompt. That would make your acknowledgement longer than your assignment. Focus on the substantive ones that directly contributed to your work.

Include:

Don’t include:

When I’m running the current iteration of GradeMap (a Claude skill I built) against my own real QUT assignments, the same distinction applies. The substantive conversations about rubric interpretation or assignment structure are the ones worth documenting. The quick “how do I format this citation” lookups aren’t.

The Academic Integrity Line

Here’s what trips up most students: what’s the difference between legitimate AI assistance and academic misconduct?

Generally acceptable:

Generally not acceptable:

The key is that AI should enhance your learning, not replace it. You’re still doing the thinking, researching, and writing. The AI is just helping you understand and organise your thoughts. Using AI for university study without cheating covers this distinction in more detail, and why AI detection tools are dying explains why transparency beats evasion.

When Tools Don’t Require Citation

Not every AI interaction needs acknowledgement. Most universities don’t require you to cite:

Again, check your own institution’s current guidance. Some universities are tightening the rules on coaching tools too.

The Over-Acknowledgement Safety Net

When in doubt, acknowledge more rather than less.

If you’re genuinely unsure whether an AI interaction needs acknowledgement, include it. A slightly longer methodology note is infinitely better than an academic integrity investigation.

Practical Templates for Common Scenarios

For concept explanation:

“I used [AI tool name and version] on [date] to help understand [specific topic] covered in Week [X] lectures. My prompt was ‘[specific question]’. I used this explanation to [specific purpose] and verified the concepts against [course readings/lectures].”

For brainstorming:

“I used [AI tool name and version] on [date] to generate initial ideas for this assignment. My prompt was ‘[question]’. I developed these initial ideas further through my own research using [sources].”

For structure feedback:

“I used [AI tool name and version] on [date] to get feedback on my assignment structure. I provided my outline and asked ‘[specific question about structure]’. I used this feedback to [specific changes made] while ensuring all content remained my own work.”

Remember, the goal isn’t to write the perfect acknowledgement. It’s to be transparent about your process and demonstrate academic integrity. Your marker wants to see that you’ve learned from the assignment, not just completed it.

References

University of Queensland. (2026). Guide to acknowledging and referencing AI. UQ Library Guides.

University of Sydney. (2026). Acknowledging AI tools and technologies. Academic Integrity Guidelines.

TEQSA. (2026). Artificial intelligence: advice for students. Tertiary Education Quality and Standards Agency.

American Psychological Association. (2023). How to cite ChatGPT. APA Style Blog.

Can I use AI to help write my acknowledgement of AI use?

Yes, but acknowledge it. If you use AI to help draft your acknowledgement statement, include that in the acknowledgement itself. It’s recursive but honest.

What if I forgot to keep track of all my AI conversations?

Most AI platforms (ChatGPT, Claude, and so on) keep conversation histories. Go back and review what you actually used in your assignment. For future assignments, keep notes as you go. It’s much easier than reconstructing later.

Do I need to cite AI even if I didn’t end up using any of its suggestions?

Generally no. If you asked AI for ideas but didn’t incorporate any of them into your final work, most universities don’t require acknowledgement. However, if the AI helped you understand a concept that informed your thinking, that’s worth acknowledging.