Scope of This Policy
This policy governs the use of generative AI and large language models (LLMs) as writing and preparation tools by authors, reviewers, and editors. It is distinct from the journal’s AI/ML Research Reporting Standards, which govern how authors report on AI/ML systems that are themselves the subject of a submitted study.
AI and Authorship
AI systems cannot be listed as authors or co-authors and cannot be assigned CRediT roles. Authors remain fully accountable for all aspects of the published work, including any content produced with AI assistance.
Permitted Uses by Authors
Authors may use AI tools for appropriate assistance such as language improvement, readability, formatting support, or code debugging. Other uses may be acceptable when transparently disclosed and independently validated by the authors.
Prohibited or Unacceptable Uses
- Fabricating or falsifying data, results, figures, or images using generative AI.
- Generating citations, quotations, or factual claims without independent verification.
- Using AI outputs as a substitute for scientific judgment, ethical review, or author accountability.
- Submitting AI-generated material that materially misrepresents the authors’ own work or source attribution.
Disclosure
Where AI use is material to preparing the manuscript, authors disclose the tool name, version where available, purpose, and the relevant stage of use in the Cover Letter and manuscript disclosures. Where AI contributes to analysis or code unrelated to the study’s own AI/ML methodology, the Methods section describes the role sufficiently for reproducibility.
Reviewer and Editor Use
Reviewers and editors must not upload confidential manuscripts, data, or peer-review correspondence to external AI systems unless the journal has expressly permitted the use and adequate confidentiality safeguards are in place. Generative AI must not be used to create the substance of a peer-review report or editorial assessment.
AI Detection
AI-detection tools, where used, are screening aids only. A detector score does not by itself constitute evidence of misconduct. Concerns about undisclosed AI use are evaluated through editorial inquiry and appropriate evidence.