AI in academia: what universities, journals and funders allow (2026)

This page is for students, researchers and supervisors who need a quick, sourced answer on what is permitted. We checked each policy on its own page in November 2026 and update this post every quarter. It is part of our series on the structure of a research paper; the practical workflow is in how to write a research paper with AI.

What do journals and publishers allow?

Every major publisher we checked permits AI assistance with human oversight, requires disclosure of substantive use, and bars AI authorship. They differ on where the disclosure goes and what counts as exempt.

An AI use policy is a publisher's or institution's written rule on when and how authors, reviewers and editors may use AI tools, and how that use must be reported.

Elsevier

Elsevier allows authors to "use AI tools to support them," but says these tools "must never be used as a substitute for human critical thinking, expertise and evaluation." Disclosure goes "in a separate AI declaration statement" placed "immediately before the references," and it appears in the published article. Authors "should not list AI tools as an author or co-author, nor cite AI tools as an author." AI must not be used "to create or alter images that represent primary observed or experimental data" [1].

Springer Nature

"Large Language Models (LLMs), such as ChatGPT, do not currently satisfy our authorship criteria," because authorship carries accountability. LLM use "should be properly documented in the Methods section (and if a Methods section is not available, in a suitable alternative part) of the manuscript" [2]. Its researcher guidance tells peer reviewers not to "upload manuscript content to unsecured or public AI systems" [3].

Wiley

Wiley says AI tools "cannot fulfil the role of, nor be listed as, an author of an article," and that "authors must take full responsibility for the accuracy of all content." Disclose in the acknowledgments when AI helps with drafting, editing, translation or formatting, and in figure captions for AI-made visuals. "AI-edited photographs are not permitted in Wiley journals" [4].

IEEE

IEEE says "the use of content generated by artificial intelligence (AI) in an article... shall be disclosed in the acknowledgments section." The AI system and the sections that use AI-generated content must be identified. Editing and grammar help is "generally outside the intent" of the policy: "disclosure as noted above is not required, but recommended" [5].

APA

APA journals do not allow AI as an author and hold authors responsible for verifying any information and citations an AI tool provides. Disclose in the most relevant part of the manuscript, often the introduction or method, and keep the AI output in case the editor asks for it [6].

ICMJE and COPE

These bodies set norms that many journals adopt. The ICMJE says chatbots "should not be listed as authors" because they "cannot be responsible for the accuracy, integrity, and originality of the work," that authors must describe AI use in the cover letter and the manuscript, and that "referencing AI-generated material as the primary source is not acceptable" [7].

COPE's position statement (February 2023) says AI tools cannot meet the requirements for authorship, that authors who use AI in writing, images or data work must disclose it in the methods or a similar section, and that authors are fully responsible for the content [8].

What do research funders allow?

Funders allow AI in preparing applications but hold applicants to originality and accuracy. All four we checked restrict reviewers.

Details, quotes and what they mean for a grant are in our AI grant writing guide.

How do universities regulate AI in academia?

Universities set rules for students and staff, and they vary more than journals. Most policies fall into one of three patterns, often with each instructor choosing for their own course:

1. Not allowed: AI tools may not be used for assessed work, sometimes except for spelling and grammar. Use is treated as an academic integrity breach.
2. Allowed with disclosure: AI may be used for named tasks (brainstorming, search, language editing) if the student says how. Drafting the assessed text is often excluded.
3. Allowed, with responsibility: AI use is open, but the student is responsible for accuracy and must cite or acknowledge it.

Theses usually follow the graduate school's policy and the examiners' expectations, not the course's. Ask your supervisor which applies and get it in writing before you rely on AI for any chapter.

Use this order to find your rule: the assignment or course syllabus first, then your department's guidance, then the university's academic integrity policy. If they disagree, the most specific rule normally applies, but ask.

What counts as authorship when AI is involved?

An author is a person who made a substantial contribution to the work, helped draft or revise it, approved the final version and agrees to be accountable for it. That is the ICMJE's standard, and it is why AI cannot be an author: it cannot approve a version or answer for errors [7].

For AI in academia, then, the question is not "did AI write any words?" but "who is accountable for every word?" If AI drafted a paragraph you did not check, you are accountable for an error you never saw. Practical lines most policies draw:

How to write an AI disclosure statement

An AI disclosure statement names the tool, the task, the sections affected, and confirms human review and responsibility. Adapt one of these templates to your journal's placement.

Declaration section (Elsevier style) [1]:

During the preparation of this work, the author(s) used [tool name] in order to [reason]. After using this tool, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.

Acknowledgments (IEEE or Wiley style):

The authors used [tool name, version or date] to [draft early versions of Sections 1 and 4 from the authors' notes / edit language in the discussion]. All AI-generated text was reviewed and revised by the authors, who take full responsibility for the content.

Methods (Springer Nature or APA style, for analysis use):

We used [tool name, version] to [write R code for the mixed-effects models / screen titles and abstracts as a second reviewer]. All outputs were checked by [author initials]; screening disagreements were resolved by discussion.

How to keep a record of your AI use

Keep evidence as you go, not after a question is raised. APA asks authors to keep AI output in case an editor requests it [6].

1. Save prompts and outputs for any AI-drafted section, with dates.
2. Keep your search log: databases, terms, dates and counts.
3. Keep version history, so you can show how a draft changed.
4. Note which sections used AI, and how, while you write. Your disclosure statement comes from this note.
5. Keep your data and analysis files separate from any AI-drafted text.

Keep a record in AI for Research

AI for Research keeps some of this record for you. It does not decide what your institution allows.

• The editor keeps version history for every document, and History saves every AI task with its inputs and result, so you can show what a tool produced and when.
• The authorship report records time spent writing and how much text was typed, pasted and accepted from AI, and shares it as a link anyone can verify. It is evidence of how a document grew, not proof, and it is not an AI detector. See the authorship report.
• Citation check confirms each reference exists and flags retractions (free), and Claim check shows whether cited studies support a claim (2 credits). See the citation checker.
• Projects can be shared with a supervisor as "can comment," so they see the work as it develops.

The free plan has 45 credits a month and the authorship report is part of the editor; see pricing.

FAQ

Can you use ChatGPT in research papers?

Usually yes, as an assistant, if you disclose substantive use and stay responsible for the content. Elsevier, Wiley, IEEE, Springer Nature, APA and ICMJE journals allow AI help with disclosure, and those that address authorship all rule out AI authors. You cannot use it to fabricate data, images of results or references. Your university or course may ban some uses, so check those rules too.

Do journals allow AI?

Most major journals allow AI-assisted writing with disclosure, ban AI authorship, and restrict AI-generated images. Elsevier wants a separate declaration before the references, IEEE wants it in the acknowledgments, and Springer Nature wants it in the methods. Reviewers at most publishers must not upload manuscripts into AI tools. Always read your target journal's own author guidelines, because individual journals can be stricter.

Can AI be listed as an author on a paper?

No. COPE, ICMJE, Elsevier, Springer Nature, Wiley and APA all say AI tools cannot be authors. Authorship means being accountable for the work, approving the final version and answering questions about it, and an AI tool can do none of these. Describe AI use in the disclosure statement or methods instead, and do not cite an AI tool as an author.

What should an AI disclosure statement include?

Name the tool and, ideally, the version or date used. Say what it did, such as literature search, drafting the introduction or editing language, and which sections it affected. State that the authors reviewed and edited the output and take full responsibility for the content. Put it where the journal asks: a declaration section, the acknowledgments or the methods.

Does NIH allow AI in grant applications?

NIH does not ban AI tools, but notice NOT-OD-25-132 (July 2025) says NIH will not consider applications substantially developed by AI, or with sections substantially developed by AI, to be the applicant's original ideas, and it caps applications at six per principal investigator per calendar year. NIH reviewers have been barred from using generative AI on applications since 2023 under NOT-OD-23-149.

Next steps

• Put the rules into practice: how to write a research paper with AI.
• See what happens when the rules are broken: AI-generated research papers.
• Writing a grant? Read AI grant writing.
• Back to the hub: the structure of a research paper.

Sources

1. Elsevier. Generative AI policies for journals. https://www.elsevier.com/about/policies-and-standards/generative-ai-policies-for-journals (checked November 2026)
2. Springer Nature. Editorial policies: Artificial Intelligence (AI). https://www.springernature.com/gp/policies/editorial-policies (checked November 2026)
3. Springer Nature. AI guidance for our researchers and communities. https://group.springernature.com/gp/group/ai/ai-guidance-for-our-researchers-and-communities (checked November 2026)
4. Wiley. Using AI tools in your research. https://www.wiley.com/en-us/publish/article/ai-guidelines/ (checked November 2026)
5. IEEE Author Center. Submission policies: Guidelines for Artificial Intelligence (AI)-Generated Text. https://conferences.ieeeauthorcenter.ieee.org/author-ethics/guidelines-and-policies/submission-policies/ (checked November 2026)
6. American Psychological Association. APA Journals policy on generative AI. https://www.apa.org/pubs/journals/resources/publishing-tips/policy-generative-ai (checked November 2026)
7. International Committee of Medical Journal Editors. Recommendations, Section V.A: Use of AI by authors. https://www.icmje.org/recommendations/browse/artificial-intelligence/ai-use-by-authors.html (checked November 2026)
8. Committee on Publication Ethics. Authorship and AI tools: COPE position statement, 13 February 2023. https://publicationethics.org/guidance/cope-position/authorship-and-ai-tools (checked November 2026)
9. National Institutes of Health. NOT-OD-25-132: Supporting Fairness and Originality in NIH Research Applications. 17 July 2025. https://grants.nih.gov/grants/guide/notice-files/NOT-OD-25-132.html (checked November 2026)
10. National Institutes of Health. NOT-OD-23-149: The Use of Generative Artificial Intelligence Technologies is Prohibited for the NIH Peer Review Process. 23 June 2023. https://grants.nih.gov/grants/guide/notice-files/NOT-OD-23-149.html (checked November 2026)
11. National Science Foundation. Notice to Research Community: Use of Generative Artificial Intelligence Technology in the NSF Merit Review Process. 14 December 2023. https://www.nsf.gov/news/notice-to-the-research-community-on-ai (checked November 2026)
12. European Research Council. Current position of the ERC Scientific Council on AI. 1 December 2023. https://erc.europa.eu/news-events/news/current-position-erc-scientific-council-ai (checked November 2026)
13. European Research Council. ERC clarifies limits on AI use in grant evaluation. 24 March 2026. https://erc.europa.eu/news-events/news/erc-clarifies-limits-ai-use-grant-evaluation (checked November 2026)
14. UK Research and Innovation. Use of generative artificial intelligence in application preparation and assessment. 23 September 2024. https://www.ukri.org/publications/generative-artificial-intelligence-in-application-and-assessment-policy/use-of-generative-artificial-intelligence-in-application-preparation-and-assessment/ (checked November 2026)