CMPUS
Alex
Leo
Sofia
Maya
Jordan
Lexi
Quinn
Owen
Tara
Max
Sia
Jade
Noah
Ashley
Nova
KaiUsing AI at university is cheating when it breaks your course's rules or when you submit AI work as your own. Rules are set course by course, and at schools like the University of Toronto, AI is not allowed unless your instructor permits it. Check the syllabus and assignment instructions, ask when unsure, disclose and cite any AI use, verify everything, and keep your drafts.
Is using AI cheating at university?
Using AI is cheating when it breaks the rules for that course or assignment, or when you hand in AI-generated work as if you wrote it. The same use can be fine in one course and an academic offence in the next.
In a 2026 KPMG Canada survey, 75 per cent of Canadian students said they use generative AI in their schoolwork, but only 53 per cent said they had received formal instruction on using it responsibly. Students and educators also disagree: 72 per cent of educators said using generative AI to assist with schoolwork is cheating, compared with 50 per cent of students.
Use keeps rising. KPMG's 2025 survey found 73 per cent of Canadian students used generative AI for schoolwork, up from 59 per cent in 2024 and 52 per cent in 2023. The real question is whether your use fits your course's rules.
Who decides what AI use is allowed?
Your instructor does, through the syllabus and the instructions for each assignment. Defaults differ between universities, and some are strict.
The University of Toronto's academic integrity guidance says it plainly: "The University of Toronto takes the stance that students are not allowed to use generative AI in a course unless the instructor explicitly permits it." The same page warns that one instructor may set different AI rules in different courses, and that using AI when it is not permitted is treated as an academic offence. Toronto Metropolitan University sets a similar default: no generative AI for coursework unless your instructor says otherwise.
Check your own university's policy page, then find the rule for your specific course:
- Read the syllabus. Look for a section on generative AI, authorized aids or academic integrity.
- Read the assignment instructions. U of T notes these are often more detailed than the syllabus, and the rules can change from one assessment to the next.
- Ask before you start. If anything is unclear, email your instructor or TA and keep the reply.
- Never carry permission across courses. A yes in one course says nothing about another, even with the same instructor.
Example question to your instructor. "For the research essay, may I use an AI tool to explain concepts from the readings and to suggest grammar fixes on my draft? I would not use it to write any part of the essay. If that is allowed, how would you like me to disclose it?"
What AI use is usually fine, and what is not?
AI that helps you learn is lower risk than AI that produces the work you hand in. Your course rules always win, and in a course that bans AI completely, even the first rows may be off limits.
| Use | Signal | Why |
|---|---|---|
| Explaining a lecture concept or reading in simpler words | Usually fine | It works like a tutor. You still do the thinking and the writing. |
| Quizzing yourself on your own notes before a test | Usually fine | It is practice, not work you submit. |
| Making a study schedule or breaking an assignment into steps | Usually fine | You are planning your time, not producing content. |
| Brainstorming topics or getting feedback on your thesis or outline | Ask first | It shapes graded work, so many instructors want it disclosed. |
| Grammar fixes, paraphrasing tools or AI rewriting features | Ask first | U of T warns that some courses do not permit paraphrasing apps or Grammarly's AI writing partner. |
| Finding sources or translating your own writing | Ask first | If allowed, read and cite the original sources, and disclose any AI translation. |
| Pasting AI-written sentences into your work as your own | Almost always misconduct | The APA Style team calls this academic misconduct for students. |
| Using AI on a quiz, test or take-home exam without permission | Almost always misconduct | It is an unauthorized aid. |
| Running AI text through a rewriter to hide where it came from | Almost always misconduct | It is still AI work presented as yours. |
How do you disclose and cite AI use?
If your instructor gives a disclosure format, use it. If not, cite the AI in your course's citation style and add a short note saying what you used it for.
The University of Toronto Libraries advises that when AI is permitted, you should cite any content it produced and any functional use, such as editing or translating. Some instructors also want your prompts and the AI's outputs in an appendix, as U of T's guidance notes.
APA style
APA's 2025 guidance names the company that made the tool as the author. To cite a specific chat, create a shared link and follow this pattern:
OpenAI. (2026, October 14). Feedback on my outline about urban heat islands [Generative AI chat]. ChatGPT. [shared chat URL]
In the text, write (OpenAI, 2026). If you only used AI to edit or refine your own writing, APA says you can cite the tool in general instead: OpenAI. (2026). ChatGPT [Large language model]. https://chatgpt.com
MLA style
The MLA Style Center does not treat an AI tool as an author. The title describes what was generated, usually by quoting your prompt:
"Explain the causes of the 1837 Rebellions in Upper Canada" prompt. ChatGPT, model [name], OpenAI, 14 Oct. 2026, [shared chat URL].
In the text, use a shortened prompt, such as ("Explain the causes"). MLA also recommends acknowledging substantive AI use beyond the citation, for example in a note, because a citation alone does not show where the AI's input ends and yours begins.
Example disclosure note. I used ChatGPT (OpenAI, 2026) to explain two concepts from the course readings and to suggest grammar fixes on my final draft. I wrote all of the text, accepted or rejected each edit myself, and checked every source against the original.
How do you show your work if anyone asks?
Keep a simple record of how each assignment came together. It is your best evidence if your work is ever questioned.
- Write in an app that keeps version history, or save a dated copy of each draft.
- Keep your reading notes, outlines and rough brainstorms.
- Save a list of every source you opened, with links.
- Save shared links to any AI chats you used, or export the prompts and outputs.
- Keep any email where your instructor or TA said what was allowed.
These match what U of T Mississauga suggests students bring to an academic integrity meeting: drafts or previous versions of the assignment, lecture notes, and copies of the sources used.
Why do you need to check everything AI tells you?
AI tools can state false things with confidence, including sources that do not exist, and you are responsible for everything you submit.
In a study published in Scientific Reports, researchers checked 636 citations from short literature reviews written by ChatGPT. Of the GPT-3.5 citations, 55 per cent were fabricated, compared with 18 per cent for GPT-4. Newer models may do better, but the MLA still warns that AI tools can make up sources.
- Find every reference yourself. Search for it in your library's catalogue. If you cannot find it, do not use it.
- Open and read the source. Confirm it says what the AI claimed, and quote the source, not the chat.
- Check facts against your course material. Your lectures and readings are the standard you will be marked against.
- Explain it back in your own words. If you cannot, you have not learned it yet.
Example prompts that keep AI in a helper role
The safest prompts ask AI to teach, question or plan, and say clearly what it should not do. Use them only where your course allows AI.
I'm a second-year economics student. Explain price elasticity of demand in plain language with one everyday example. Then ask me three questions to check that I understood. Do not write any part of my assignment.
Here is my essay question and my working thesis. Ask me five questions that would help me sharpen the argument. Do not rewrite the thesis for me.
Here are my lecture notes on cell signalling. Quiz me with ten short-answer questions, one at a time, and at the end tell me which topics I should practise more.
What should you do if you are accused of using AI?
Respond promptly, stay calm, and use your process record to show how you did the work. Procedures differ by university, so read your school's academic integrity process before any meeting.
- Do not ignore the message. Reply, confirm any meeting and ask what the concern is.
- Reread the rules. Check exactly what the syllabus and assignment said about AI, and find any permission you got in writing.
- Gather your evidence. Drafts, version history, notes, sources and any AI chats you disclosed.
- Be honest about your process. Explain what you did plainly. If you used AI in a way you believed was allowed, say so and show why.
- Get support early. Talk to your registrar or academic advisor, your school's academic integrity office, or your student union's advocacy service.
At U of T, for example, students can get legal advice from Downtown Legal Services at any point in the process, and UTM students may bring a support person, such as a friend, family member or student union member, to their meeting.
Where CMPUS fits
CMPUS is an AI study platform built around academic integrity. Its Classmates plan, explain, question and quiz you, and they are not designed to hand back finished assignments.
Alex works out what you are trying to do and sends you to the right specialist. Leo, the Personal Tutor, explains concepts at your level and checks that they landed. Maya, the Essay Planner, helps you build a thesis, outline and source plan, and you write the essay.
CMPUS does not change your course's rules. If your instructor bans AI, that includes CMPUS, and if they require disclosure, disclose it. Professors who want students to get AI help inside their own policy can run a course edition built on their syllabus, rubric and AI-use policy.
Not sure where to start? Alex works out what the task actually is and walks you to the right specialist.
Enter the CMPUS 5 free interactions a day. No sign-up, no card.
Questions people ask
01Is using ChatGPT for my essay cheating?
02What if my syllabus says nothing about AI?
03How do I cite ChatGPT in APA style?
04How do I cite AI in MLA style?
05Do I need to cite AI if I only used it for editing?
06What should I do if I am accused of using AI?
Sources
- Generative AI and Academic Integrity · University of Toronto Mississauga
- Generative Artificial Intelligence in Learning and Teaching at TMU · Toronto Metropolitan University
- Citing Artificial Intelligence (AI) · University of Toronto Libraries
- Citing generative AI in APA Style: Part 1, Reference formats · APA Style
- How do I cite generative AI in MLA style? (Updated and Revised) · MLA Style Center
- Beyond Citation: Describing AI Use in Your Work · MLA Style Center
- Majority of students use AI, but educators sound alarm (2026) · KPMG Canada
- Generative AI boom among Canadian students raises dilemmas (2025) · KPMG Canada
- Fabrication and errors in the bibliographic citations generated by ChatGPT · Scientific Reports (Walters and Wilder, 2023)
- Academic Integrity Meeting Overview · University of Toronto Mississauga
CMPUS guides are general advice, not your course's rules. Your syllabus and your instructor decide what is allowed. Spotted something out of date? Tell us.