It depends on your school, your program and your committee. Some graduate schools leave the decision to your advisor, some ask for approval in advance and some courses allow no AI tools at all. Read your own rules first, then ask your advisor by email and keep the answer.
No one sees their own blind spots, with or without AI help. Esmond points them out in your draft from within your field, along with factual errors and inconsistencies.
At the schools quoted here, someone close to your work. At Dartmouth's graduate school, “Programs, advisors, and course instructors have authority to allow or disallow GenAI use for specific research projects, courses, or assignments, depending on the academic objectives.” In RIT's computing PhD program, a policy applies to everyone and a committee can go further: “Advisors and dissertation committees may impose more stringent rules on Generative AI use for documents and presentations.”
So two students at one school can get different answers. Canada's University of Guelph spells out the order of events for graduate students who plan to use these tools: they “must seek unambiguous approval for planned uses in advance from their advisor(s) and advisory committee.” Ask before you upload, because a yes afterward is harder to get.
A hypothetical example of a commentYour literature review says earlier work is divided on this question, but your discussion calls the finding well established. Which of the two do you stand behind?
A first comment on one part of your own draft, right away and without an account.
Get a free previewNone of the policies quoted on this page has a separate line for AI comments on a draft. They speak of generative AI in research and in the text you hand in. You have to work out whether a tool that only comments counts, and you shouldn't assume it doesn't. Dartmouth adds that cloud AI tools should not be used to process unpublished research, so ask about the upload itself.
A policy can also decide for you when nobody has said anything. Arizona State's School of Social Transformation has this default: “In the case where an instructor or committee does not have an ai policy, or it is not clearly conveyed, it will default to the Permissible Use of Generative Ai with Disclosure policy.” At that school, if nothing is clearly conveyed, AI use is permitted with disclosure. Elsewhere it may work differently, so don't read your own program's silence as a yes.
Our page on the AI disclosure statement shows what such a line can contain.
| If your policy says | For AI feedback on a draft | What to do |
|---|---|---|
| No AI tools at all | This covers a tool that only comments | Don't use one for this work |
| AI tools with permission from your advisor or committee | Feedback from an AI tool needs that same permission | Ask first, and name the tool and the task |
| AI use must be disclosed | Comments from an AI tool are a use you can report | Add a line to your statement |
| AI-generated text is not accepted | A tool that only comments adds no text, though the policy may not say so | Ask how your program reads it |
| Nothing about AI | This settles nothing, and your academic integrity code can still apply | Ask your advisor and keep the reply |
We quote policies literally, with the source and the date we read them. They change, so open the original before you rely on it.
"Can I use AI?" gets a vague answer. A narrow question is easier to answer, and easier to rely on later. For example: "I'd like to run chapter 3 through an AI tool that comments on the content and adds no text to it. Is that fine at this stage, and how should I report it?" Then cover these points:
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.
Esmond returns comments and adds no text to your thesis. Can I use Esmond? lists what to check and how to name it in a statement, and AI thesis feedback that never writes for you shows what you get back.
No. One produces text that could go into your thesis; the other comments on text you already have. Your policy may still treat both as AI use, so check it.
That depends on your rules. Comments from a tool your program permits, reported the way it asks, are a different case from handing in text a tool produced. If your program bans all AI tools, comments fall under that ban too.
If your policy asks you to disclose AI use, assume yes unless your advisor tells you otherwise. Name the tool, what you used it for and when.
Before. Policies quoted on this page ask for permission or approval in advance, and a yes by email is easier to rely on later.
Read on November 4, 2026.
Get your paper evaluated before you submit