1. Core principles
- human authors, reviewers, and editors remain responsible and accountable for their work and decisions;
- AI output must not be treated as authoritative without verification;
- confidential manuscripts, participant data, reviewer reports, and editorial correspondence must not be uploaded to public AI services;
- AI cannot be listed as an author, reviewer, editor, source, or responsible contributor;
- AI use must comply with copyright, privacy, data protection, research ethics, and third-party terms.
3. Required disclosure
4. Images, figures, audio, and video
- generative AI must not create or alter original research images or data in a misleading manner;
- ordinary adjustments are permissible only when applied consistently and when they do not obscure, remove, or introduce information;
- AI-generated graphical abstracts, illustrations, or cover art require prior editorial approval, rights clearance, and clear disclosure;
- when AI image analysis is part of the method, authors must provide reproducible details and may be asked for original files.
5. Reviewers
Reviewers must not upload a manuscript, extract, data, figure, or review report to a public or externally retained AI system. Peer review is a human scholarly responsibility. AI must not generate the evaluation, recommendation, or confidential report.
6. Editors and journal staff
Editors and staff must not upload submissions, reports, author data, ethics records, allegations, or decision letters to public AI services. AI must not make editorial decisions, select reviewers autonomously, or replace conflict, ethics, reference, or integrity assessment.
7. Assessment of suspected AI use
AI detectors are not sufficiently reliable to prove misconduct on their own. Concerns are assessed using the manuscript, disclosures, source verification, supporting records, author response, and other evidence. Undisclosed or prohibited use may result in correction, rejection, expression of concern, or retraction depending on materiality.



