Audit Reveals Surge in Fake Citations Across Biomedical Literature
A comprehensive audit of 2.5 million biomedical papers published between January 2023 and February 2026 has uncovered a significant rise in fabricated citations. The study, published in The Lancet, analyzed 97 million references and identified nearly 3,000 papers containing fake references that could not be traced to known publications. Researchers utilized an automated pipeline involving large language models to detect mismatches between cited titles and actual records in databases like PubMed and Crossref. The findings indicate that the number of publications with fabricated citations increased twelvefold from 2023 to 2025, suggesting a rapidly growing integrity crisis in scientific publishing. Study co-author Maxim Topaz describes these figures as conservative estimates, representing only the tip of the iceberg. Experts suspect generative AI plays a significant role in this trend, although human fabrication remains possible. This issue threatens the reliability of biomedical research, as fake citations undermine the verification process essential for scientific progress. The analysis highlights the urgent need for improved detection mechanisms and stricter editorial standards to combat the contamination of academic literature with non-existent sources.
Wire timeline
Audit Reveals Surge in Fake Citations Across Biomedical Literature
A comprehensive audit of 2.5 million biomedical papers published between January 2023 and February 2026 has uncovered a significant rise in fabricated citations. The study, published in The Lancet, analyzed 97 million references and identified nearly 3,000 papers containing fake references that could not be traced to known publications. Researchers utilized an automated pipeline involving large language models to detect mismatches between cited titles and actual records in databases like PubMed and Crossref. The findings indicate that the number of publications with fabricated citations increased twelvefold from 2023 to 2025, suggesting a rapidly growing integrity crisis in scientific publishing. Study co-author Maxim Topaz describes these figures as conservative estimates, representing only the tip of the iceberg. Experts suspect generative AI plays a significant role in this trend, although human fabrication remains possible. This issue threatens the reliability of biomedical research, as fake citations undermine the verification process essential for scientific progress. The analysis highlights the urgent need for improved detection mechanisms and stricter editorial standards to combat the contamination of academic literature with non-existent sources.
Nature