A practical framework for responsible use of generative and other AI tools by authors, reviewers and editors while protecting accountability, confidentiality and research integrity.
Responsible AI
Authors remain responsible for the accuracy, originality and integrity of submitted work. AI tools cannot assume authorship responsibility or answer for the work.
Journals should specify when and how material use of AI tools must be disclosed, including the location and level of detail expected.
Editors and reviewers should not upload confidential manuscripts or identifiable unpublished content into AI services where privacy, retention or reuse cannot be appropriately controlled.
AI-generated text, references, summaries and analyses can contain errors or fabricated details. Human verification remains essential.
| Context | Potential use | Editorial safeguard |
|---|---|---|
| Manuscript preparation | Language support, summarisation, coding or drafting assistance. | Require author responsibility, verification and disclosure where policy requires it. |
| Peer review | Assistance with organisation or analysis. | Protect manuscript confidentiality and prohibit delegation of reviewer judgment to AI. |
| Editorial screening | Workflow support, metadata checks or triage assistance. | Keep accountable human oversight and test for bias/error before operational use. |
| Production | Copyediting, tagging, accessibility or metadata assistance. | Use quality control and preserve the meaning of the scholarly record. |
AI-detection outputs should not be treated as definitive proof of misconduct. Editors should consider evidence, tool limitations, manuscript context and author explanations. Similarity screening and AI detection address different questions.
Journals should define acceptable uses of generated or altered images, data and multimedia, with particular attention to research evidence, documentation, disclosure and the possibility of misleading manipulation.
AI policies should be specific enough to guide behaviour but adaptable enough to respond to rapidly changing tools. Journals should review the policy periodically and communicate major changes clearly.
Receive updates on training, conferences, editorial resources, opportunities and scholarly publishing developments.