Artificial intelligence is changing how academic publishing works, and peer review is no exception. AI tools are now being used to screen manuscripts for plagiarism, suggest reviewers, check statistical methods, and even draft review reports.
For peer reviewers, this raises practical and ethical questions that need careful thought. The key issue is not whether AI exists in the workflow, but whether it supports or replaces the reviewer’s own judgment.
Where AI is already involved in peer review
Many AI-driven checks are performed before a manuscript reaches a reviewer. Editorial systems may check for plagiarism, image issues, statistical inconsistencies, and reviewer matches. Most reviewers do not directly interact with these tools, but they are already part of the process behind the scenes.
AI becomes more visible during peer review and the writing of the peer review report itself. Tools can help summarize a manuscript, organize thoughts into a structured report, or suggest clearer phrasing. These uses vary widely, and the boundaries are not always clearly defined.
Industry guidelines about the use of AI in peer review
Guidance from the Committee on Publication Ethics emphasizes confidentiality in peer review and the reviewer’s responsibility for their reports. Unpublished manuscripts should not be shared with third parties unless journal policies explicitly allow it.
This principle extends to any external tool that processes manuscript content. Reviewers are expected to be cautious, protect confidentiality, and follow journal-specific policies where available. Reviewers also remain fully responsible for their evaluations.
How AI can help in peer review
- Using a tool to improve grammar, clarity, or flow after the review is written can be acceptable when journal policy allows it.
- AI can help find related literature that may strengthen a reviewer’s perspective.
- AI can support structure once the reviewer’s analysis is complete.
These uses should support the human effort of reviewing, not replace careful reading, independent assessment, and expert judgment.
Disclosure
If you have used AI in any part of your review, including assistance with language or structure, disclose it. Transparency is a reasonable professional norm regardless of whether a specific journal policy exists.
A note in confidential comments to the editor is usually sufficient, for example: I used a language editing tool to check the clarity of this report after completing my own evaluation. It is also helpful to clarify what tool you used and how you used it.
The bottom line
AI is a tool, and like all tools, it can be used well or badly. In peer review, the test is whether the expert judgment at the heart of the evaluation is genuinely yours.
If you have read the manuscript carefully, assessed the methods and conclusions independently, and arrived at your own evaluation, AI assistance with writing or literature search is not inherently a problem when confidentiality is protected and use is disclosed. If you have not done that work, no AI tool can substitute for it.
