A student who uses AI on coursework should first check what the course allows, then disclose the use and cite it in the style the instructor requires. The University of Toronto Mississauga (UTM) says generative AI is off limits in a course unless the instructor explicitly permits it, Stanford's Office of Community Standards treats it like help from another person, and the Modern Language Association (MLA) asks writers to cite any AI content they paraphrase, quote, or build on.
Key takeaways
- UTM's default is no generative AI unless the instructor says yes, and the rule can change from one assignment to the next.
- Stanford's guidance, adopted February 16, 2023, treats AI assistance like help from another person and tells students to disclose any use that is more than incidental.
- MLA says not to list an AI tool as an author, and to name the tool, version, publisher, and date the content was generated.
- Where a rule is unclear, both institutions point students to the instructor, and asking before submitting is the safest step.
Two institutions, two different defaults
Universities do not share one rule about AI, and the first thing to learn is which default a campus uses. UTM's academic integrity page for students states that students are not allowed to use generative AI in a course unless the instructor explicitly permits it. If use is permitted, the page tells students to follow that instructor's guidelines, not to assume the same rules carry over to other courses taught by the same person, and to read both the syllabus and each assignment's instructions, because assignment rules on authorized aids often differ. It also advises checking before using paraphrasing apps or Grammarly's AI writing partner, since some courses do not permit them. Using AI where it is not permitted counts as an academic offence and can lead to an allegation of misconduct.
Stanford took a different route. Its generative AI policy guidance, adopted on February 16, 2023, says that using or consulting a generative AI tool is treated analogously to assistance from another person. Using one to substantially complete an assignment or exam, for example by entering the questions into it, is not permitted. Instructors may allow or disallow some or all uses in a given course, and students who are unsure are told to ask.
What to disclose
Stanford's guidance tells students to acknowledge generative AI use other than incidental use and to default to disclosing assistance when in doubt. It does not set a citation format. The UTM page is less specific: it does not impose a universal requirement, but notes that an instructor who permits AI may ask students to cite the portions the tool generated and to place the prompts and outputs in an appendix.
The University of Toronto's central student page on AI adds a simple instruction: a student who is unsure whether AI is allowed in an assignment should consult the instructor or graduate supervisor for clarification. In both cases the wording of the course instructions governs, which is why a screenshot or saved copy of the instructor's reply is worth keeping.
How MLA asks for a citation
The MLA's post on citing generative AI was published March 17, 2023, and notes that MLA released an updated, revised version of its guidance in August 2025, so the current wording should be checked before relying on any single example. The principles it sets out are consistent. Cite a tool whenever text, an image, data, or other material it created is paraphrased, quoted, or incorporated. Acknowledge functional uses, such as editing prose or translating words, in a note, in the text, or elsewhere suitable. Vet any secondary sources the tool mentions, since it does not always cite sources or show what a source contributed. MLA does not recommend listing the tool as the author.
The entry maps onto MLA's core elements. The title of the source describes what the tool generated and includes the prompt if it is not already in the text. The container is the tool's name, italicized. The version is given as specifically as possible, followed by the publisher, which is the company that made the tool, and the date the content was generated, and finally a general URL. MLA's own example, for a paraphrased ChatGPT answer about the green light in The Great Gatsby, lists the prompt as the title, then ChatGPT, the February version, OpenAI, the March 8, 2023 generation date, and the tool's URL.
Side by side
| Source | Default stance | Disclosure or citation |
|---|---|---|
| UTM academic integrity page | Not allowed unless the instructor explicitly permits it | May need to cite generated portions and add prompts and outputs in an appendix |
| Stanford Office of Community Standards (February 16, 2023) | Treated like help from another person; no substantial completion of work | Acknowledge more than incidental use; disclose when in doubt; no format set |
| MLA (March 17, 2023; revised August 2025) | A style guide, so it sets no permission rule | Cite content used; acknowledge functional uses; do not list the tool as author |
APA and Chicago publish their own guidance on citing AI, and secondary guides indicate that it has been revised since 2023. The official pages could not be opened for this article, so students using those styles should read the current edition or the library guide their course points to.
A disclosure note, illustrative example
The note below is an illustrative example of the sort of statement an instructor who permits AI might accept. It is not required wording, and the instructor's template always takes precedence.
"For this essay, ChatGPT was used on October 3, 2026, to suggest three possible outlines for the introduction. The outline in section two was adapted from its second suggestion. All sources were found and read by the author, and the prompts and responses are attached in Appendix A."
A short check before submitting
- Is generative AI permitted in this course and on this assignment, according to the syllabus and assignment sheet?
- Does the permission cover grammar or paraphrasing tools as well as chatbots?
- Is every AI-generated passage, image, or data point marked and cited in the required style?
- Are the prompts, tool name, version, and date saved for an appendix or note?
- Has every source the tool mentioned been located and read, not just copied?
This article summarizes published policies from three organizations as of October 2026. It is not academic advice, and the rules of a specific school, program, or instructor decide what is allowed.
FAQ
Can an AI tool be listed as an author?
MLA does not recommend it. The tool is named as the container of the generated content, and the company that made it is the publisher.
What if a syllabus says nothing about AI?
At UTM the stated default is that it is not allowed unless the instructor says so. Stanford treats it as help from another person, and both institutions tell students to ask the instructor.
Does light use such as checking grammar need disclosure?
It depends on the course. Stanford asks for acknowledgment of use beyond the incidental, and UTM tells students to check with the instructor about paraphrasing and writing-assistant tools. MLA lists editing and translating among the functional uses that should be acknowledged.
Is a citation enough to make AI use acceptable?
No. A citation shows what was used, but permission comes from the course rules.
Related reading
- what a 2026 survey says Canadian students do with AI
- trust between students and teachers when AI is in the room
- how many teachers and students use AI
- accessibility rules for AI tools
More in AI Basics.
Last verified: October 2026
Sources
- Generative AI and Academic Integrity, University of Toronto Mississauga, accessed October 2026
- Citing generative AI in MLA style (MLA Style Center post), MLA Style Center, March 17, 2023, with note on the August 2025 revision
- Generative AI Policy Guidance, Office of Community Standards, Stanford University, adopted February 16, 2023
- AI education and use of GenAI for students, University of Toronto, accessed October 2026