In this round-up, we look at four recent discussions shaping the conversation around peer review, research integrity, and the growing role of AI in scholarly publishing. From rethinking how research is evaluated to tightening safeguards against manipulation and setting clearer boundaries for AI use, these developments reflect a publishing ecosystem adapting to rapid change.
Rethinking the peer-review gate
In his guest post on The Scholarly Kitchen, Dmitry Kochetkov asks whether peer review needs an evolution or a revolution. He points to growing submission volumes, reviewer shortages, lengthy review cycles, and the additional pressure created by AI-assisted publishing. The author argues that the traditional accept-or-reject model can make research dissemination unnecessarily slow and may disadvantage novel or interdisciplinary work. Kochetkov explores the Publish-Review-Curate (PRC) model, where research can become publicly available while evaluation continues through transparent, ongoing review. He also highlights the potential of publication-level “trust markers” to help readers assess the credibility of individual research outputs. Read the full article here.
When peer review itself becomes the problem
A report from Retraction Watch highlights the consequences of weaknesses in peer-review safeguards. Elsevier’s International Journal of Biological Macromolecules has retracted 120 papers following an investigation into systematic manipulation of peer review in guest-edited special issues. The investigation also identified concerns involving plagiarism, image duplication, and authorship. The case is part of a wider pattern of integrity concerns surrounding special issues and comes as the journal’s publication volume has grown sharply in recent years. Elsevier has said it has tightened its checks on Guest Editors to identify potential problems earlier. The scale of the retractions underscores the importance of robust editorial oversight as publishers manage increasingly large volumes of research. Read the full article here.
Keeping human judgment at the center of AI-assisted review
In a commentary published in Management Science, Tinglong Dai responds to proposals for using AI to manage rising submission volumes. While agreeing that journals need clear governance around AI, Dai questions whether increasing submissions should be viewed as an existential threat to peer review. He argues that human judgment remains particularly important when assessing novelty and contribution, noting that AI systems trained on existing patterns may struggle to identify genuinely new ideas. The commentary supports AI as a non-final input into peer review, while keeping humans responsible for the judgments that shape scholarly contribution. Read the full article here.
JAMA sets clearer boundaries for AI use
In their editorial for JAMA, Annette Flanagin, Roy H. Perlis, and Kirsten Bibbins-Domingo outline updated guidance for authors using AI in medical publishing. The guidance allows AI for activities such as research, manuscript preparation, translation, and data visualization, provided its use is appropriately disclosed and authors remain responsible for the content. At the same time, JAMA prohibits AI use in peer review and in drafting Opinion manuscripts, Letters to the Editor, and online Comments. It also advises against using AI to generate or manage references because of the risk of fabricated citations and restricts AI-generated or manipulated clinical images and multimedia. The updated guidance reflects a broader effort to balance the efficiencies of AI with the need for transparency, confidentiality, and accountability. Read the full article here.
If you’ve come across a piece lately that sparked reflection or raised important questions, feel free to share it with the ReviewerOne community.
