An AI-assisted audit conducted by the Columbia University School of Nursing has uncovered nearly 3,000 peer-reviewed medical papers containing fake citations that do not exist in scientific databases. This study, published in The Lancet on May 7, 2026, analyzed 2.5 million biomedical papers published between January 1, 2023, and February 18, 2026. The research team identified 4,046 fake citations across 2,810 papers, with the rate of such citations increasing more than 12-fold since 2023, particularly from mid-2024, coinciding with the rise of AI writing tools. This issue poses a significant risk to patient care, as medical professionals may unknowingly rely on fabricated references when making treatment decisions. Maxim Topaz, PhD, who led the study, emphasized the importance of verifying references during the submission process. The authors recommend that publishers implement rigorous verification measures and that research integrity databases track fake references systematically. Alarmingly, 98.4% of the affected papers had not undergone any publisher action at the time of the audit, highlighting the urgent need for accountability in academic publishing.