For EECS students, yes, on terms your advisor and committee set. The School of Electrical Engineering and Computer Science published an AI policy in June 2026 that covers theses, dissertations and other documents submitted to examination committees. On November 9, 2026 we found no Graduate School or university-wide rule for AI in theses, so this page covers the EECS text only. Where no norms have been agreed, grammar checking, search and autocomplete need no approval, and any substantive use needs prior approval. Ask your advisor first, and disclose substantive use in the thesis.
If your AI statement clearly doesn't cover what you describe elsewhere in your thesis, Esmond points it out.
The EECS document is called Guidance on Drafting AI Use Policies. Most of it addresses courses, and one section, AI Use in Research Work, applies to research by EECS graduate students, undergraduate researchers and faculty. It names theses, dissertations, manuscripts, grant proposals, technical reports and documents submitted to examination committees. The version history lists June 2026 as the initial version.
Your major advisor sets the norms for research under their supervision. For a thesis the policy goes further: “For thesis and dissertation work, AI use expectations should be set with the full committee and documented early, such as at the program of study meeting or sooner.”
If that conversation has not happened yet, a default applies: “Where no norms have been set, AI tools may be used for grammar checking, search, and autocomplete without prior approval; any substantive use requires prior approval.” So ask your advisor first before you use Esmond, or any AI tool beyond those three, for your thesis or dissertation.
The policy calls itself a floor. Rules of venues, sponsors, funders or an IRB come first where they are stricter. EECS also says a college-wide AI policy for the College of Engineering is in development, and that OSU policy governs where the documents conflict.
A hypothetical example of a commentYour acknowledgments thank a tool for comments on chapter 2, but Figure 5.3 shows synthetic benchmark traces and the methods chapter never says how they were generated. Who produced that data, and is it described where a reader would look?
A first comment on one part of your own draft, right away and without an account.
Get a free previewAppendix B sorts research uses into three tiers, adapted from Resnik and Hosseini (2026). The research section says it lists which uses are substantive and which need disclosure, but the tiers themselves are labelled by disclosure only. It does not replace the approval your committee gives.
Feedback on your own work sits in the middle tier. The list of optional disclosures includes this item: “Validate or give feedback on your own existing ideas, text, or code.” It is not in the tier that needs no disclosure, so it is more than grammar or search. Whether it counts as substantive use under the default rule is something the text does not settle, which is one more reason to raise it with your advisor.
| Tier | Examples in Appendix B | What it means for a thesis |
|---|---|---|
| Disclosure mandatory | Hypotheses or experiment design, drafting or significant revision, translation, data analysis, synthetic data or figures | Name the tool and its role in the document |
| Disclosure optional | Finding references, organizing a paper, feedback on your own ideas, text or code | Follow what your committee agreed |
| No disclosure | Grammar and spelling, search, clarity of a single sentence, autocomplete | Nothing to add, unless your committee asks |
Venue rules come first. Where they say nothing, EECS asks you to disclose substantive use in the methods section, the acknowledgments or an author contribution statement, and to name the tool, the type of task and its scope. Image editing and image generation get their own mention. The policy says the same expectation applies to documents submitted to examination committees, which covers your thesis drafts and the final version.
For a thesis with chapters that will also go to a conference or journal, check each venue's rules early. Several major venues named in the policy require disclosure of substantive AI use, and some prohibit specific uses.
EECS treats the researcher as the author and as accountable for the result. Every factual claim, citation, quantitative result and code path has to be checked independently, because tools invent references and misstate prior work. The research question, design choices, analysis decisions and interpretation have to be your own. You must be able to explain any text, analysis or code to your advisor, committee or a reviewer.
The policy recommends a similarity check on all documents submitted to examination committees, with a university-provided tool or an equivalent. If you and your advisor disagree about AI use, EECS says to take it to the Associate Head of Graduate Programs.
The research section also limits what you upload. Sensitive or confidential material should not go into non-approved AI tools. The policy lists proprietary or sponsor-restricted material, identifiable information about human subjects, and other people's unpublished work used without their permission, such as a coauthor's draft or a collaborator's code. Sponsor agreements and IRB protocols can restrict AI use further.
For a thesis built from joint papers, this matters. Before a chapter with a coauthor's unpublished text or a lab partner's code goes into any tool, get their permission, and check whether a sponsor agreement covers the project. Whether your own unpublished draft counts as sensitive or confidential material is not settled by the EECS text, so ask your advisor before you upload it to Esmond or any tool Oregon State has not approved.
This page describes one school. Other Oregon State units may have their own expectations or none in writing. To find yours:
This example is made up. A PhD student in computer science never set AI norms with his committee. He asks a tool for comments on his related-work chapter and thanks it in the acknowledgments. He also has a tool generate synthetic network traces for two figures in chapter 5. Under Appendix B the comments are an optional disclosure and the synthetic data is a mandatory one, and under the default rule both may count as substantive use that needed prior approval. His document discloses the lighter use and is silent on the one the policy weighs more.
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 evaluates a draft from within your field and points out factual errors, inconsistencies and blind spots, which puts it close to the feedback item in Appendix B. See how Esmond works and a sample report. Under the EECS default, substantive use needs prior approval, so ask your advisor first before you use Esmond for this thesis or dissertation, and agree the terms with your committee. If you go ahead, disclose it as agreed, with the tool, the task and which chapters it covered. Don't upload a chapter with a coauthor's unpublished work without their permission, and ask your advisor whether your own draft counts as confidential material before you upload it. See also whether you can use Esmond. Oregon State has not reviewed or approved Esmond.
Not in the EECS policy. It lets the advisor and committee set norms, and without norms it allows grammar, search and autocomplete freely and requires prior approval for any substantive use.
We found none on November 9, 2026. This page covers the School of EECS. Other units may have their own statements.
Appendix B lists feedback on your own text as an optional disclosure. Your committee can ask for more, so follow what you agreed with them.
The policy says students should escalate disagreements to the Associate Head of Graduate Programs.
Read on November 9, 2026.
Get your paper evaluated before you submit