Citation Verification as a Standard Component of Peer Review: Reflections and Recommendations in the Era of AI-Generated Scholarship

Peer review is essential to maintaining the integrity of the scholarly record, yet reference lists often receive limited scrutiny. As generative artificial intelligence (AI) becomes increasingly integrated into scientific writing, this oversight poses growing risks. Large language models can generate citations that appear legitimate but are partially or entirely fabricated, allowing unverifiable references to enter published literature. Such citations may evade plagiarism detection software and standard editorial safeguards. Drawing on one reviewer’s experience and literature on peer review and research integrity, this article argues that fabricated references constitute falsification, undermiing reproducibility, distorting evidence hierarchies, and compromising scientific integrity. It highlights gaps in existing peer-review guidelines, particularly the absence of explicit expectations and tools for citation verification, and proposes practical, low-burden recommendations for reviewers and journals. Integrating routine citation spot-checking into peer review represents a feasible, scalable approach to preventing AI-generated citation fabrication and strengthening accountability in scientific publishing.