Generative AI: disclosed when used, prohibited where harmful.

Banning AI tools entirely is unrealistic and unproductive. Allowing them without disclosure is dangerous. Our policy strikes a middle path: AI is allowed for certain tasks, prohibited for others, and must always be disclosed.

For Authors

Permitted AI Use (with disclosure)

• Language editing and grammar checking (e.g., Grammarly, ChatGPT for proofreading).
• Translation between languages.
• Code review and debugging for computational research.
• Literature search assistance (with manual verification of every citation).
• Generating non-substantive content like abstracts (with substantial human revision).

Prohibited AI Use

• Generating original research findings or analyses.
• Writing the methods, results, or discussion sections wholesale.
• Fabricating data, references, or citations.
• Listing AI as an author or co-author (AI cannot be an author).
• Generating images, figures, or charts that are presented as actual experimental output.
• Using AI to evade plagiarism detection.

Disclosure Requirement

All permitted AI use must be disclosed in a dedicated "AI Use Declaration" section before the references. The declaration must specify:

• Which AI tool was used.
• For which purpose (language editing / translation / code review / etc.).
• In which sections of the manuscript.
• How outputs were verified.

For Reviewers

Prohibited AI Use

Reviewers may NOT use AI tools to generate review content. The reviewer's role is to provide expert human judgment. Submitting AI-generated reviews is a serious violation of the reviewer code of conduct and grounds for removal from the reviewer pool.

Permitted AI Use

Reviewers may use AI tools for personal language assistance (improving the clarity of their own writing) and for translation (if reviewing in a non-native language). All AI use must be at the reviewer's personal level —never sending manuscript content to public AI tools (this violates confidentiality).

For Editors

Editors may not use public AI tools to assess manuscripts, as this violates author confidentiality. Internal AI-screening tools (iThenticate's AI detection, GPTZero, etc.) used as part of the editorial workflow are permitted.

AI Screening at Submission

Every submitted manuscript is screened by AI-detection tools (GPTZero, Originality.ai, Copyleaks) in addition to plagiarism checking. The AI-detection score is part of the editor's triage information. Manuscripts flagged as high-probability AI-generated are returned to authors with a request to declare and justify AI use.