Why Disclosure Rules Matter Now
A disclosure policy typically requires authors to state where and how generative AI was used, whether for drafting, editing, analysis, coding, or image creation, while making clear that the human author remains fully responsible for accuracy, originality, and ethical compliance. It usually prohibits listing AI as an author, demands verification of all outputs, and asks for enough detail to let editors and readers assess influence on the work, including prompts or workflows when relevant.
Also worth reading: How Should AI Authorship Disclosure Contracts Protect Creators and Clients? · Does Amazon KDP require AI disclosure for books published in 2026? · How Can Publishers Use an AI Disclosure Policy Template to Build Reader Trust?
Such policies also expect disclosure of the specific tool, version, and purpose, plus confirmation that no confidential or protected data was improperly shared. They may require authors to describe human oversight, cite AI-assisted content appropriately, and update disclosures if use changes during revision. The goal is not to ban assistance but to preserve trust, reproducibility, and accountability across scholarly publishing workflows for readers, editors, and institutions.
What Counts as AI Authorship
A disclosure policy usually requires authors to state clearly when and how generative AI was used in preparing a manuscript, including drafting, editing, analysis, or image creation. It asks for enough detail to let readers and editors judge whether the work remains human-directed and whether the methods are reproducible. The policy also typically requires authors to carefully verify all AI-assisted content, correct errors, and ensure that citations, data, and claims are accurate before submission.
It generally does not treat AI as an author because software cannot fully take responsibility for the work. Instead, human authors remain accountable for the full manuscript, including any AI-generated text or visuals. Disclosure may need to appear in acknowledgments, methods, or a dedicated statement, depending on the publisher. Clear reporting helps distinguish legitimate assistance from undisclosed automation and supports trust in scholarly and professional publishing.
Journal Policies and Practical Steps
An AI authorship disclosure policy requires three core commitments: defining what AI tools may legitimately do in the writing process, specifying where and how that use must be declared, and clarifying that AI cannot hold authorship itself. Taylor & Francis's generative AI policy exemplifies this structure, permitting AI for language polishing and drafting assistance while requiring authors to disclose its use in a dedicated statement. The European Journal of International Law took a similar route, publishing its first policy on generative AI to establish transparency expectations before problems arose rather than after.
For journals and publishers, the practical steps are straightforward. Draft a policy that distinguishes AI-assisted editing from AI-generated content, since the latter raises deeper questions of accountability and verification. Require a disclosure statement at submission, and train editors to evaluate it consistently. Early-career authors, particularly in fields like medicine, benefit from explicit supervision of AI use so that disclosure becomes a form of professional formation rather than a compliance checkbox. Finally, revisit the policy annually, because tool capabilities and community norms are shifting faster than most editorial boards anticipate.
Risks of Incomplete AI Disclosure
An AI authorship disclosure policy requires authors to state clearly when generative tools contributed to drafting, editing, analysis, or visual creation, while retaining human accountability for claims and methods. It usually asks for tool name, version, purpose, and extent of assistance, plus confirmation that authors reviewed outputs for accuracy, bias, and originality. Policies also prohibit listing AI as an author because responsibility cannot be assigned to software, and they may require disclosure in acknowledgments, methods, or a dedicated statement. The goal is transparency rather than punishment, helping editors and readers assess how the work was produced.
Such policies often distinguish between minor language polishing and substantive intellectual contribution, with stricter rules for the latter. They may require disclosure of prompts, data handling, privacy safeguards, and whether AI-generated text or images were independently verified. Institutions and journals increasingly expect authors to document human oversight, cite sources AI suggested, and avoid uploading confidential material into public tools. A complete disclosure therefore functions as an audit trail: it shows where machine assistance ended and human judgment began, preserving trust in the scholarly record.
Building a Compliant Workflow
An AI authorship disclosure policy requires authors to state when and how generative AI tools were used in manuscript preparation. It prohibits listing AI as an author, since systems cannot assume responsibility for accuracy or integrity. Authors remain accountable for all submitted work. Policies typically demand disclosure statements in acknowledgments or methods sections, naming the tool, version, and purpose. Transparency lets reviewers and readers assess potential bias or error. Failure to disclose may bring retractions or sanctions. Ultimately, the policy preserves trust in scholarly communication while permitting responsible, documented use of emerging technologies.
Requirements vary across publishers, but common elements include describing AI involvement, confirming human verification of outputs, and ensuring no confidential material was improperly processed. Authors must also disclose AI use in peer review or editorial tasks when applicable. Institutions increasingly align internal guidelines with publisher standards. Compliance involves maintaining records of prompts and outputs for audit purposes. As policies evolve, authors should consult current journal instructions before submission. A compliant workflow integrates disclosure checkpoints, reducing omission risk. Clear documentation supports reproducibility and accountability.
AI Disclosure Policy Comparison
| Requirement | What It Means | Typical Policy Language |
|---|---|---|
| Disclose AI use | Name tools and scope of assistance | "Authors must state any generative AI use" |
| Retain human accountability | Authors verify and own all content | "The author remains responsible for the work" |
| No AI authorship | AI cannot be credited as an author | "AI tools are not eligible for authorship" |
| Transparent methods | Explain prompts, edits, or analysis | "Describe AI assistance in the methods section" |