When you agree to peer review a manuscript, you accept a specific and serious obligation of confidentiality. The manuscript has not been published, and its content, data, methods, and arguments are entrusted to you as part of a privileged process. Although many reviewers recognize this obligation in principle, applying it in real situations can be more complex than the simple rule suggests. This post sets out what the confidentiality obligation in peer review covers, where the gray areas lie, and what the real-world consequences of a breach look like.

Why confidentiality matters in peer review

The obligation of confidentiality in peer review rests on a straightforward rationale: authors submit unpublished work at a point of genuine intellectual vulnerability. They are sharing findings, methods, and data that have not yet been validated or deemed ready for publication. If any part of this information is disclosed, whether through unintentional carelessness, curiosity, or deliberate exploitation, the harm can be significant.

The most serious form of breach of confidentiality is the exploitation of a manuscript’s ideas or data for the reviewer’s own research. This could manifest as publishing findings that draw on unpublished work encountered during peer review or citing ideas from a manuscript as if they originated independently. This is both an ethics violation and, in some cases, a form of academic theft. The Ethical Guidelines for Peer Reviewers outlined by the Committee on Publication Ethics (COPE) are explicit about this point: reviewers must not use ideas, data, or methods from a manuscript in peer review for their own purposes before the work is published.

A less dramatic but serious breach is simply discussing the manuscript or its contents with others. An author presenting at a conference does not expect to find a member of the audience questioning them from a copy of their unpublished paper, (see examples from COPE case studies below). The damage this could cause to the authors’ intellectual contribution and to the concept of trust in the peer review system is real, regardless of whether any deliberate misconduct was intended.

What confidentiality in peer review covers

The confidentiality obligation in peer review applies to:

• The existence of the manuscript: You cannot tell others that you are reviewing a particular piece of work, not even in general terms that allow identification of the manuscript’s content or its authors.
• The content: You need to maintain the confidentiality of the manuscript’s methods, data, results, arguments, and conclusions. This also includes details shared in correspondence with the editor about the manuscript.
• Your review: The contents of your peer review report remain confidential unless the journal operates an open peer review model where reviews are published alongside the manuscript. Even where your identity as a reviewer is known, the content of the report is not yours to share freely.
• Post-publication: Confidentiality does not end when the paper is published. You should not describe or discuss the private peer review correspondence, including what you recommended, what the authors said in their responses to reviewers, or what other reviewers shared, unless the journal explicitly publishes this information as part of an open review model.

In journals operating open peer review models, where reviews are published alongside the article, the confidentiality norms are explicitly different. In some models, your identity as a reviewer is known to the authors from the outset; in others, your full review is published with the draft. One rule remains constant: your need to uphold confidentiality of the research is always applicable. You should not discuss the manuscript publicly before the journal publishes both the paper and the reviews.

What you cannot do

• Share the manuscript, or any part of it, with a colleague, student, or collaborator without first obtaining the editor’s permission.
• Cite, reference, or build upon unpublished findings from the manuscript in your own work before it is published.
• Discuss the manuscript’s existence or its contents on social media, at conferences, or in any public or private forum outside the journal peer review workflow.
• Retain copies of the manuscript after the peer review is complete. Many journals explicitly ask reviewers to delete or destroy copies.
• Use the peer review process strategically to delay a competitor’s publication while advancing your own work in the same area.

The grey areas

In some cases, complex circumstances make it difficult for reviewers to understand the need for confidentiality.

• Co-reviewing: Sometimes, reviewers enlist the support of a junior, peer, or senior colleague for the peer review, either as a collaborative/learning exercise or because a section falls outside their own expertise. While this practice is not frowned upon, reviewers must ensure that they involve co-reviewers only after obtaining explicit written permission from the editor. The co-reviewer’s name and relevant details must also be provided to the journal. Not only does this align with the essential principle of transparency, but it also upholds the spirit of academic collaboration and helps ensure that the co-reviewer also gets credited for the review. Passing a manuscript to a co-reviewer without permission is a breach of confidentiality, regardless of intention. It also implies means the co-reviewer has taken on a confidentiality obligation they were never formally made aware of.

• Seeing a conference presentation of material under review: If you’re reviewing a manuscript and attend a presentation at a conference where the same or closely related findings are described, you are in an awkward position. You cannot discuss the manuscript with the presenter; you also cannot use your knowledge of the paper in your assessment of the presentation. If the work appears to have been disclosed in a way that concerns you, for example, the presentation describes findings identical to unpublished data in the manuscript, this may be worth flagging with the editor. But your role is to note the observation, not to investigate, discuss, or share.

Reviewing a manuscript that closely resembles one you have peer reviewed before: If you are invited to review a manuscript that resembles work you encountered while reviewing for another journal, the confidentiality obligation applies in both directions. You cannot draw on the previous manuscript in your review. You can, however, note to the editor that you have seen related unpublished work elsewhere, without identifying the source, this may provide relevant information for the editor while respecting your earlier obligation.

What happens when confidentiality is breached

COPE cases illustrate the range of consequences that follow a breach of confidentiality. In one case, a reviewer forwarded a manuscript to a trainee (not a content expert) without authorization. Following an investigation, the reviewer was removed from the journal’s reviewer list and the institution was contacted to inform them about the breach. In another case, a reviewer used material from a manuscript under review in a conference presentation. The authors recognized their own unpublished data and informed the editor about the potential breach in confidentiality. The reviewer was barred from reviewing for the journal for three years, and the case was referred to the reviewer’s institution.

In the most serious scenarios, where a breach involves the deliberate exploitation of unpublished ideas, the consequences can extend to formal misconduct proceedings. Even where the intent was not malicious, consequences such as being removed from a journal’s reviewer panel and having an institution informed of the circumstances is a significant professional outcome.

For more discussions, resources, and perspectives on peer review, join the ReviewerOne Community.

References

• COPE ethical guidelines for peer reviewers
• COPE case 14-06: Possible breach of reviewer confidentiality
• COPE case study: More than a breach of confidentiality
• COPE discussion document: Who owns peer reviews?
• COPE case 11-29: Reviewer asks trainee to review manuscript