As Peer Review Advances: Auditing the Gap Between Procedural Performance and Research Value

Declaration of AI-Assisted Research Large language models (LLMs) were substantially used to assist with English drafting and translation, manuscript organization, identification of literature and policy sources, preliminary documentary coding, and the preparation of analytical figures and supporting code. The central research questions, conceptual distinctions, substantive interpretations, and decisions concerning the scope and conclusions of the study were directed and revised by the human author, including through original Chinese-language formulations. LLM-generated outputs were not treated as independently verified evidence. The human author retains responsibility for source verification, critical assessment of generated material, and approval of the final manuscript. No AI system is designated as an author.