Only with prior permission. The Faculty of Graduate & Postdoctoral Studies (GPS) guidelines, dated February 2, 2024 and updated December 13, 2024, say graduate students must seek prior permission and approval from their supervisor and supervisory committee members before using generative AI in thesis research and writing. Any use then goes in the preface of the thesis and in publications from the research. GPS treats submitting AI-generated work without saying so as cheating.
If your AI statement clearly doesn't cover what you describe elsewhere in your thesis, Esmond points it out.
The summary at the top of the GPS guidelines sets the order of events: “Graduate students who intend to use generative AI tools in their thesis research and writing must seek prior permission and approval from their supervisors and supervisory committee members.” Permission comes first, and the guidelines name both your supervisor and the members of your supervisory committee. If you are unsure whether every committee member has to agree, check with your department.
The guidelines don't say what form the permission takes. Our suggestion is to ask by email, name the tool and the task, and keep the reply, so you can show later what was agreed and when.
A hypothetical example of a commentYour preface says no AI tools were used, but your methods chapter describes an AI tool that translated your interview transcripts. Which statement is accurate?
A first comment on one part of your own draft, right away and without an account.
Get a free previewThe GPS document answers separate questions for different parts of a graduate degree. This is what it says for each.
| Part of your degree | What GPS says |
|---|---|
| Thesis research and writing | Prior permission and approval from your supervisor and supervisory committee members; disclosure in the preface; misconduct involving AI in the thesis or publications goes through the GPS academic misconduct disciplinary process |
| Candidacy proposals, qualifying exams, comprehensive exams | Answered under the same question as the thesis, with the same prior permission; misconduct involving AI in comprehensive exams goes through the GPS academic misconduct disciplinary process |
| Course papers, project reports, assignments | Follow the instructor's statement of expectations in the syllabus or on eClass; if there is none, ask the instructor for permission first |
| Literature review, data analysis, simulation | Follow your program's discipline-specific regulations if it has them, and get prior permission from your supervisor and committee |
| Manuscripts for journals and conferences | Consult the journal's or conference's own AI policy and author guidelines, and disclose the use in publications from your research |
| Awards and scholarship applications | The same GPS guidelines apply, and you should stay informed of changing funding agency expectations |
On placement GPS is specific: “You must disclose your use of generative AI tools in the preface to your thesis.” Depending on the norms of your discipline, a transparency statement can also appear in the methodology or materials and methods section, the acknowledgments, or an appendix. The preface statement is still required.
The transparency statement must explain how AI algorithms, tools, or applications were used in your research and writing process. GPS says you must acknowledge and cite each tool you used in a well-known citation style, and points to the APA guide as an example. Your program and faculty may set further disclosure requirements, so check with your department before you finalize the preface.
GPS says graduate programs and units may develop subject or discipline-specific guidelines, and that you must comply with them. Those rules may cover when AI is appropriate, or not, in data collection, analysis, modeling, simulation, visualization, or academic writing. The guidelines present these as requirements in addition to the GPS rules: the prior permission from your supervisor and committee still applies.
GPS links its guidelines to the Student Academic Integrity Policy, which defines plagiarism as representing the work of another individual or other source as your own. On AI it adds one sentence: “Submitting work created by generative AI and not indicating such would constitute cheating as defined above.”
The guidelines also cite the GPS regulations on originality: a master's thesis should be an original contribution, and a doctoral thesis must embody the results of original investigations and analysis. As a graduate student, GPS says, you are responsible for the ethical and responsible use of AI tools in your thesis.
GPS warns that generative AI tools are known to fabricate citations and references. Before you use a source a tool recommends, the guidelines say you should verify its reliability, authenticity, and accuracy in scholarly databases, and they name the University of Alberta Library, Scopus, Web of Science, Google Scholar, and WorldCat as examples. They also note that output can be biased, discriminatory, incomplete, or inaccurate, for instance references recommended only by citation counts that are gender-biased, culturally insensitive, or geographically inequitable.
The guidelines name data privacy and confidentiality as two key concerns. AI tools may collect and share personal data and the research data you upload, and output may reveal confidential information from your research. In the GPS text, being aware of a tool's privacy and confidentiality policies comes before its use. If your thesis involves interview data or other confidential material, raise that when you ask your supervisor and committee for permission (our suggestion).
This example is made up. A doctoral student in rehabilitation science gets approval from his supervisor and two committee members to use an AI tool for transcribing pilot interviews. A year later he wants to use a different tool to suggest a coding frame for the full data set. The guidelines don't say whether one approval covers a later, different use, so the safe reading is that he asks again before he starts. He also checks whether his program has discipline-specific rules on AI in analysis, and lists both tools in the preface.
Esmond evaluates every assignment before you submit. It never writes it.
Esmond is an AI tool that doesn't write for you. If your course doesn't allow any AI tools, don't use Esmond for that work. If you're unsure, ask your instructor or advisor.
The GPS guidelines don't mention feedback tools. They ask for prior permission before you make use of any generative AI tools for the thesis, so the safe reading is that AI comments on a draft chapter fall under that rule. Ask your supervisor and committee members first, and if they agree, include the use in your preface. GPS wants you to be aware of a tool's privacy and confidentiality policies before use, so read the Privacy Notice. The University of Alberta has not reviewed or approved Esmond.
Esmond reads the draft you wrote and returns comments on factual errors, inconsistencies, and blind spots, next to the passages they are about. It adds no sentences to your thesis. To show your committee what the comments look like, see how Esmond works and open the sample report.
For thesis research and writing, GPS says you need prior permission and approval from your supervisor and your supervisory committee members. For a graduate course, follow the instructor's statement, or ask the instructor if there is none.
Yes. The guidelines answer the question for candidacy proposals, qualifying exams, comprehensive exams, the thesis, and publications together, and say misconduct involving AI in comprehensive exams, the thesis, or publications is handled through the GPS academic misconduct disciplinary process.
In the preface. Depending on your discipline, a transparency statement can also go in the methods section, the acknowledgments, or an appendix. Your program and faculty may have more requirements.
GPS says submitting work created by generative AI without indicating it would constitute cheating under the Student Academic Integrity Policy. Talk to your supervisor before you submit.
Read on November 7, 2026.
Get your paper evaluated before you submit