Reviewers remain fully responsible for the content, accuracy, fairness, and confidentiality of their review reports, regardless of whether any AI-assisted tools are used.

Reviewers must not upload, paste, submit, or otherwise transfer the manuscript, reviewer materials, unpublished data, tables, figures, source code, supplementary files, or confidential editorial correspondence into publicly available Generative AI systems or third-party AI services.

This restriction applies because manuscripts under peer review may contain unpublished intellectual property, confidential research findings, personal information, proprietary datasets, or other information that has not yet been made publicly available.

Where permitted by the journal and where no confidential manuscript content is disclosed, reviewers may use AI tools for limited administrative or language-related assistance, such as:

• Improving general grammar or language clarity.
• Checking general readability of a reviewer's own text.
• Assisting with non-confidential writing or formatting tasks.
• Helping organize the reviewer's own ideas before preparing the final report.

Generative AI must not replace the reviewer's own expert assessment. Reviewers are expected to personally evaluate the manuscript's originality, methodology, data, analysis, results, conclusions, and scientific contribution.

AI-generated or AI-assisted text may contain inaccurate information, fabricated references, inappropriate recommendations, or misleading interpretations. Reviewers must independently verify any information generated or suggested by an AI system before incorporating it into a review report.

• Uploading the complete manuscript to a public AI chatbot.
• Uploading confidential sections of the manuscript to an external AI service.
• Using AI to generate the substantive review without meaningful human evaluation.
• Allowing AI to determine the final recommendation independently.
• Using unpublished information obtained through peer review for AI-assisted research.
• Disclosing the identity of authors, reviewers, editors, or confidential editorial information to an AI system.

The reviewer remains responsible for the final review report. The use of AI assistance does not transfer responsibility for errors, inappropriate comments, confidentiality breaches, or ethical violations to the AI tool.

Generative AI tools must not be used as an independent basis for accepting, rejecting, or recommending revision of a manuscript. Editorial decisions must remain under the responsibility of qualified human editors.