Written by human beings but with a grammar-check programme keeping an eye on these non-native speakers’ expressions, this editorial introduces EJIL’s first Policy on Use of Generative Artificial Intelligence (Gen AI), a form of artificial intelligence that some authors use to analyse data and to create content, including text or figures. The policy was adopted after extensive discussion in EJIL’s Advisory and Executive Boards. It applies to EJIL’s entire ecosystem: journal articles, book reviews, blog contributions, social media posts and yes, also podcasts. This editorial focuses on submissions to EJIL. The policy aims to provide clarity ex ante on which uses of Gen AI are permissible and impermissible, and which uses need to be disclosed and which need not (for instance, using Gen AI to check grammar, spelling and punctuation).
The rapid integration of Gen AI into academic research and writing has prompted important questions about authorship, intellectual contribution and research ethics. The three pillars of EJIL’s Gen AI policy are disclosure, author responsibility and accountability. These pillars ensure that editors, reviewers and readers can assess the role of Gen AI and protect the integrity of the scholarly process.
The Continuing Need for Human Intelligence in Scholarly Production
EJIL’s policy begins from the premise that while Gen AI may complement, it should not replace, human intelligence in scholarly production. Only humans can be authors, given the need for accountability and trust in research. The policy distinguishes between permissible and impermissible uses of Gen AI by analogy to legal-academic best practice on authorship – namely on the question whether and in what way the named author(s) of a submission are expected to acknowledge the contributions of others to the submission.
The key question for deciding whether Gen AI use is permissible or not is whether a submission retains a significant intellectual contribution from the named author(s), with the author’s contribution clearly outweighing that of Gen AI. The policy prohibits the use of Gen AI that would render the author’s contribution negligible or that would, if made by a human, require co-authorship. Conversely, in case the contribution, if made by a human, would require acknowledgement, the use of Gen AI is permitted – provided the author properly discloses and explains such use.
Disclosure of Gen AI
Given the policy’s starting point of the intellectual contribution of humans, authors must disclose uses of Gen AI beyond de minimis exceptions, such as checking grammar and spelling. Disclosure of Gen AI use does not preclude the submission from being considered by EJIL; it allows the editors to assess whether it is compatible with the policy and how it is best disclosed to readers.
Authors who have used Gen AI must disclose this through the form titled ‘Declaration Regarding Use of Generative Artificial Intelligence in the Research and/or Writing of EJIL Submissions’, which is part of the submission of manuscripts through ScholarOne. The form requires authors to disclose details on the particular Gen AI used; the precise purposes of such use; how the Gen AI was used; and the measures the author took to verify the accuracy of the output of the Gen AI. Authors who have not used Gen AI (or whose use falls below the de minimis threshold) need not submit the form. If authors remain unsure whether specific uses of AI require disclosure after carefully considering EJIL’s Gen AI policy, they should contact EJIL’s Managing Editor.
Responsibility of (human) authors
EJIL’s Gen AI policy also has a strong focus on the responsibility of (human) authors. Authors remain fully responsible for the accuracy, integrity and originality of their entire manuscript, also if the use of Gen AI was disclosed and permissible under the policy. They bear, for instance, full responsibility if the use of Gen AI results in plagiarism or infringement of third-party rights. Implementation and Enforcement
EJIL’s editorial team assesses the declarations of Gen AI use on a case-by-case basis. If the team discovers evidence of non-permitted Gen AI use, or of use that was improperly disclosed, it may exercise its discretion not to send a manuscript to peer review. If EJIL peer reviewers encounter material that may have been AI-generated, they should report their suspicion to the editorial team using the confidential comments box as part of their evaluation of the manuscript. The editorial team will carry out checks in such cases, including by having regard to the author’s declaration of Gen AI use. If EJIL discovers non-permitted or improperly disclosed Gen AI use after publication, or is alerted to such use by third parties, the policy foresees retraction as a possible sanction.
Gen AI and peer review
EJIL’s Gen AI policy focuses on the use of Gen AI by authors of manuscripts; it does not cover the peer review process. We summarised EJIL’s approach to the use of AI by peer reviewers in our editorial on AI: An Assistant, but No Peer, in EJIL 36 (2025) 287. In our notes to peer reviewers, we instruct them not to use AI under any circumstances. There is thus a crucial difference: reviewers cannot permissibly use any form of AI, even with disclosure; whereas authors may use AI, including Gen AI, provided they disclose such use accurately and the use is in line with EJIL’s Gen AI policy.
Adapting the AI policy
As AI continues to develop, so too must the policies that govern its use. EJIL’s approach recognises this fluidity. We welcome comments on the policy, including on how to adjust the policy as AI evolves.
As this editorial explained, EJIL’s Gen AI policy does not impose a ban on AI in research and writing. It aims to regulate the use of AI by authors with standards rather than prescriptive rules, drawing analogies to best practice in research. This also means that the policy is designed to evolve considering this practice and as AI changes.
We will keep our AI policy under review and will update it considering experience across EJIL’s entire ecosystem. We are deeply grateful to our Associate Editor Abhimanyu George Jain for taking the lead in drafting the policy and to the other Associate Editors and the Advisory and Executive Boards for productive discussions. We invite you to read our two-page Gen AI policy, whether you are thinking about submitting to EJIL or not. To be continued!
https://www.ejiltalk.org/ejils-first-policy-on-use-of-generative-artificial-intelligence/
