Ethical Considerations for Using AI during Peer Review

Artificial intelligence (AI) is going to change the future of medical publishing. Likewise, peer review as an important component of medical publishing is going to be changed. Day by day peer reviewers may find that by using AI technologies their reviews can be more robust or evidence-based; however, the dark side of using AI in peer review is compromising the confidentiality of the manuscript’s content that they are reviewing.

Furthermore, the review content provided by AI may not necessarily come from the sole efforts of the reviewers. In the newly released ICMJE recommendations, it is advised that reviewers are required to uphold the manuscript’s confidentiality, which prevents them from uploading the manuscript to any software or AI technologies where confidentiality cannot be guaranteed. It is important for reviewers to inform journals about their use of AI technology in the review process and describe how it is being utilized. Reviewers should be mindful that AI has the potential to generate output that may sound authoritative but can be inaccurate, incomplete, or influenced by biases.