How Do You Find Out Whether Your University Has an AI Policy for Researchers?

Where Is Your University’s AI Policy Hiding?

You search your university website for "AI policy" and find nothing. Does that mean researchers are free to use AI however they choose?

Probably not. Universities govern research through many overlapping policies, and AI requirements may be scattered across research integrity, data protection, information security, research ethics, intellectual property, acceptable-use, procurement, and academic-integrity documents. Some institutions have created dedicated generative AI policies for research. Others have incorporated AI into existing rules or issued guidance rather than a standalone policy.

The challenge is therefore not simply finding a document with "AI" in its title. You need to identify the institutional requirements that actually govern your proposed research activity and confirm that the information is current.

Do Not Stop After Searching for “AI Policy”

How to Locate the AI Rules That Actually Apply

Start With Official University Sources

A general web search is useful for discovery, but the final authority should normally be an official institutional source. Search the university's main website, research office, research-integrity office, library, information technology service, data-protection or privacy office, and research ethics pages.

Useful searches include the university name together with terms such as "generative AI research," "AI research policy," "AI researchers," "ChatGPT research," "AI data security," and "AI research ethics."

Do not assume that the first search result is current. Universities may retain old announcements, archived guidance, workshop materials, or pages written for students completing coursework rather than researchers conducting studies.

Look Beyond Documents Explicitly Called AI Policies

This is one of the most important parts of the search. AI governance may be distributed across existing institutional policies.

The University of Oxford provides a useful current example. It has a dedicated Policy for using Generative AI in Research, but that policy explicitly sits alongside other university policies covering research integrity, research ethics, research data management, intellectual property, data protection, cyber security, publication and authorship, export control, and open research. Finding the dedicated AI document is therefore only the beginning of the policy check.

Other institutions organise governance differently. The University of Cambridge's current Research Data Management Policy Framework, for example, explicitly discusses generative AI while stating that, at present, there is no distinct university policy for using generative AI tools in research. It directs researchers to comply with relevant university, funder, and publisher requirements.

These examples show why identifying all the policy layers governing AI use matters more than locating a document with a particular title.

Make Sure the Policy Is Actually About Research

Universities may maintain separate AI guidance for students, teaching staff, administrative employees, researchers, and technology procurement. A policy about AI in examinations does not necessarily establish what a faculty researcher may do with interview transcripts. Likewise, staff guidance for drafting routine emails may not determine whether AI can be used for research analysis.

Check the document's scope. Look for statements identifying who is covered, which activities are included, and whether the policy applies to staff researchers, postgraduate researchers, visiting researchers, contractors, students, or research conducted using university facilities.

Oxford's current research policy, for example, explicitly defines its scope to include researchers and professional staff supporting research and identifies substantive uses across the research lifecycle. That level of specificity is considerably more useful than assuming that every institutional AI document governs every activity.

Check Whether Your University Classifies Data by Risk

For many research activities, the decisive institutional question is not "Can I use AI?" but "Can I put this information into this AI system?"

Some universities classify information according to sensitivity and approve particular systems for particular data categories. Harvard University, for example, advises users not to enter confidential information, including non-public research data, into publicly available generative AI tools. Its current guidance states that confidential data may be entered only into generative AI tools assessed and approved for the relevant level of use by its information-security and data-privacy functions.

This distinction is consequential. A researcher might be permitted to use an AI service for public literature while being prohibited from uploading unpublished participant data to the public version of the same service.

Check the Date, Version, and Responsible Office

AI policies can change unusually quickly. When a policy page provides a publication date, revision date, version number, review schedule, or policy owner, record it.

Oxford's generative AI research policy illustrates why this matters. Its September 2026 version identifies a version number, amendment history, responsible service, and a termly review timetable. A screenshot or recollection from an earlier version may therefore no longer describe the current requirements.

For long projects, finding the policy once is not enough. Researchers may need a process for keeping track of changing AI requirements as the study progresses.

Check Local Rules After Finding the Central University Position

A university-wide policy may establish a baseline rather than the complete rule set. Faculties, schools, departments, laboratories, clinical units, research centres, libraries, or IT services may issue additional instructions.

Harvard's university-level generative AI guidance, for example, explicitly tells community members to review local School or Unit policies. Harvard Medical School separately maintains research-focused AI resources that bring together institutional requirements, publishing guidance, funder requirements, and information about approved tools.

This is why finding a central policy does not eliminate the need to determine whether your department or research unit has additional AI rules.

Ask the Office That Owns the Relevant Risk

If published guidance does not resolve your question, contact the office responsible for the issue rather than sending a vague question to whichever university email address appears first.

Ask a specific question. "Can researchers use AI?" invites an answer so broad that it may be practically useless. Describe the tool, task, type of information involved, and intended research purpose.

Keep Evidence of the Policy You Relied On

If AI use materially affects your research, retain enough information to establish what guidance was applicable when you made the decision. Depending on institutional practice, this might include the policy URL, title, version or revision date, relevant correspondence, ethics amendment, approval of a particular platform, or project documentation.

This does not require building a bureaucratic museum of every webpage you have ever visited. The objective is traceability for consequential decisions, especially when policies are frequently revised.

Searching for a Policy When No “Research AI Policy” Appears

Common Mistakes When Looking for University AI Rules

Use a Policy Search That Follows the Research Activity

The most efficient approach is not to collect every institutional document containing the letters "AI." Begin with the research activity you intend to perform and follow the risks it creates.

Remember that institutional permission is still only one part of the larger governance picture. Funders may impose their own AI requirements, and a journal may apply publication rules even when your university permits the tool.

Before Concluding That Your University Allows the AI Use

Frequently Asked Questions About Finding University AI Policies

Finding the Policy Means Looking Beyond Its Title

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.