When AI Reviews Train AI Reviewers: Scientific-Judgment Collapse and Mitigation

Large language models are increasingly used as automated reviewers and as assistants to human reviewers in scientific evaluation, according to a study on recursive AI peer review. As model-generated reviews enter public data and future training corpora, the research warns that later reviewers can learn from judgments produced by earlier models, a dynamic the study calls scientific-judgment collapse. The work examines mitigation approaches for this recursive feedback loop.

Large language models LLMs increasingly participate in scientific evaluation, both as automated reviewers and as assistants to human reviewers. As model-generated reviews enter public data and future training corpora, AI peer review can become recursive: later reviewers learn from judgments produc