Large Language Models and the Future of Peer-Reviewed Medical Publishing: Challenges, Editorial Responses, and a Framework for Responsible Integration

Abstract

The public release of large language models (LLMs) in late 2022 has fundamentally altered the landscape of scholarly medical publishing. LLMs now permeate every stage of the academic publishing pipeline, from manuscript drafting and peer review to editorial decision-making, with evidence suggesting that at least 13.5% of biomedical abstracts published in 2024 showed detectable LLM involvement. This rapid adoption has intersected with pre-existing structural vulnerabilities, including escalating article processing charges, publish-or-perish incentives, a chronic peer reviewer shortage, and inadequate editorial resources, creating interconnected challenges that affect all stakeholders. The response from the scholarly publishing community has been substantial but fragmented. International standards bodies, such as the International Committee of Medical Journal Editors (ICMJE), Committee on Publication Ethics (COPE), World Association of Medical Editors (WAME), have established consensus principles prohibiting AI authorship and requiring disclosure, yet individual journals range from restrictive to actively encouraging in their AI policies, and detection-based enforcement approaches have proven fundamentally unreliable, with documented biases against non-native English speakers. This narrative review characterizes the structural challenges of LLM adoption in medical publishing, provides a systematic comparison of editorial policies across major medical journals and publishers, and examines the limitations of current detection and enforcement mechanisms. Rather than focusing on prohibition and policing, the review proposes a forward-looking framework centered on three concrete proposals: 1) a three-tiered disclosure system (assistive, augmentative, substantive) integrated with the CRediT taxonomy; 2) a four-stage artificial intelligence co-editor model for deploying LLM-based tools within editorial workflows under transparent governance principles; and 3) equity-conscious policy design with international coordination through a proposed global editorial summit. The framework aims to shift the editorial paradigm from reactive enforcement to proactive governance, addressing the operational reality that editors are overburdened and underresourced for the expanding expectations placed upon them.