AI Surge Exposes 3,000 Fake Biomedical Paper Citations
An examination of 2.5 million biomedical research papers has uncovered that almost 3,000 included fraudulent citations that are not present in scientific databases, underscoring a concerning trend in academic publishing amid the increasing utilization of artificial intelligence. The papers were published from January 1, 2023, to February 18, 2026, in PubMed Central’s Open Access database, overseen by the National Institutes of Health in the United States. The findings indicated that among 97.1 million verified references, researchers identified 4,046 fake citations across 2,810 papers. This represents a more than 12-fold increase since 2023, with the most significant rise commencing in mid-2024, aligning with the emergence of AI writing tools.
This discovery has a direct impact on patients, as medical professionals base their treatment decisions on clinical guidelines,” stated lead researcher Maxim Topaz. A medical professional or clinical guideline developer lacks the means to ascertain that the evidence upon which they depend is non-existent. For instance, one study we examined contained 18 out of 30 fabricated references. “Some of those citations are already being cited by other papers and appear in systematic reviews that inform clinical care,” Topaz stated. The authors stated, “Among 97.1 million verified references, we identified 4,046 fabricated references across 2,810 papers.”
The fabrication rate increased more than 12 times, from approximately four per 10,000 papers in 2023, to 51.3 per 10,000 papers in the fourth quarter of 2025, reaching 56.9 per 10,000 papers in early 2026. The researchers advised publishers to confirm references with every paper submission and suggested that indexing services incorporate metadata into records, enabling users to evaluate the accuracy of references. They also urged major research integrity databases to establish a dedicated category for fraudulent references to enable systematic tracking and accountability. The team urged publishers to conduct retroactive screenings of current publications and to issue corrections or retractions in instances where fraudulent references undermine a paper’s conclusions.
