Gen AI Policy

  1. Brief Summary

This policy sets out the rules for the use of Generative AI tools (e.g., large language models, image generators, code generators, etc.) in the preparation of manuscripts and related materials submitted to [Journal Name]. Purpose: To maintain scientific integrity, transparency, accountability, and reproducibility while enabling the ethical use of new tools.

 

  1. Scope

Applies to all manuscript submitters (authors, co-authors), reviewers, and editors who use Gen-AI technology to: write manuscripts, formulate abstracts, prepare literature reviews, generate graphs/illustrations, generate or modify code, analyse data, or perform other tasks that impact scientific content.

 

  1. Definitions
  • Generative AI (Gen-AI): A model or tool (software/service) that generates text, images, code, sound, or other data based on training on a large corpus (e.g., LLM, diffusion models, code models).
  • Material use: any Gen-AI output included in the manuscript or supporting materials (figures, tables, code, datasets, etc.).
  • Disclosure: a formal statement in the manuscript explaining how Gen-AI was used.

 

  1. General principles
  • Transparency: All uses of Gen-AI that influence the content must be clearly disclosed in the manuscript.
  • Human accountability: Humans (authors) are fully responsible for the content, correctness, interpretation, and ethics of the work; Gen-AI cannot be claimed as author or co-author.
  • Authenticity & originality: Gen-AI output does not replace the author's original intellectual work; plagiarism of Gen-AI output remains a violation.
  • Reproducibility: Authors should provide guidelines for reproducing the use of Gen-AI (e.g., prompt, model version, parameters) when relevant for scientific validity.
  • Legal & Ethical Compliance: All uses must comply with copyright laws, personal data protection, and research ethics requirements (e.g., informed consent when output involves personal data).

 

  1. Disclosure of Gen-AI Use (mandatory)
  • Authors must include a section titled “Generative AI Use Statement” (or a section in Methods/Acknowledgements) that describes:
  • The name/type of the Gen-AI tool (e.g., ChatGPT vX, DALL·E, Stable Diffusion, Codex), including the provider (OpenAI, Hugging Face, etc.).
  • The model version, date of access, and key parameters, if available.
  • Details of how the Gen-AI output was used (e.g., drafting initial drafts, correcting language, creating illustrations, generating analysis code).
  • Who edited/reviewed the Gen-AI output, and what verification steps were taken?
  • A short example text is provided at the end of the document.

 

  1. Authorship and Contributions
  • Gen-AI should not be credited as an author. Because it does not meet the authorship criteria (responsibility, accountability, ability to approve publication).
  • If Gen-AI made substantial contributions to the manuscript, these contributions should be disclosed in the Author Contributions statement, but without naming it as an author.
  • All human authors remain accountable for all parts of the manuscript, including those generated or facilitated by Gen-AI.

 

  1. Use of specific guidelines in different parts of the manuscript
  • Text (abstract, introduction, discussion): The use of Gen-AI for language editing is permitted, but must be disclosed. Using Gen-AI to develop scientific claims, analytical results, or key interpretations is not recommended unless the output is fully verified by the authors.
  • Data analysis/code: Code generated by Gen-AI may be used, but authors must: (a) test and verify the code, (b) include the complete source code and environment (package version), and (c) explain Gen-AI's role in generating/modifying the code.
  • Images/illustrations: If used, please identify the tool, model/semantic prompt, and confirm that they do not infringe third-party copyright. If images are based on human data/identity, ensure permission or anonymisation.
  • Methods & Experiments: Methods generated by Gen-AI must be experimentally validated and transparently described.

 

  1. Peer Review and Editorial
  • Reviewers should not use Gen-AI to draft the content of a review submitted to a journal without disclosing its use. Reviewers remain responsible for their opinions and assessments.
  • Editors who use Gen-AI for editorial tasks (e.g., summarising, drafting decisions) must report such use to their fellow editors and reserve substantial decisions for full human review.

 

  1. Personal Data, Copyright, and Security
  • Do not submit sensitive personal data (PII) to public Gen-AI services unless privacy compliance has been ensured.
  • If Gen-AI output uses third-party copyrighted material (e.g., images, text), authors must provide proof of permission or ensure fair use/permission.
  • For paid or private service usage, document the data retention & deletion policies of the provider.

 

  1. Reproducibility & Documentation
  • For any use of Gen-AI that impacts the results or interpretation, authors must include (to the extent possible):
  • Service name, model version, date of access.
  • Full prompt (request text) or pseudocode of the prompt used.
  • Important parameters (temperature, max_tokens, seed) if available.
  • Generated code and environment (file requirements, library version).
  • Note: If there are security or commercial concerns, provide a summary sufficient to allow readers to evaluate the reliability of the claims.

 

  1. Plagiarism & Validity Checking
  • Manuscripts containing Gen-AI output will still be checked for plagiarism and originality. Gen-AI output that copies copyrighted text or presents false claims will be considered infringing.
  • Reviewers/editors may request verification of Gen-AI use.

 

  1. Violations & Sanctions
  • Violations of this policy (e.g., failure to disclose the use of Gen-AI, false claims, plagiarism) may result in:
  • Requests for revisions and clarifications, or
  • Rejection of the manuscript, or
  • Retraction of the publication if the violation is discovered after publication, or
  • Notification to the author's institution if the violation is serious.

 

  1. Exceptions & Special Cases
  • For editorial articles, opinion commentaries, or non-scientific materials, the policy remains in effect but can be applied more flexibly; disclosure of the use of Gen-AI is still required.
  • For datasets or code releases that use Gen-AI for data augmentation, include full documentation of the augmentation process so that other users can evaluate its impact.

 

  1. Policy Changes

The journal may update this policy as technology evolves. The version and revision date will be listed at the top.