Fake references have also surged in medical papers. It is further evidence of the rise of "paper mills," which produce fake papers for money. While fabricated studies with manipulated research data have sometimes been caught, this is the first time the extent of fake references cited in papers has been identified.
A team led by Maxim Topaz, a professor at Columbia University School of Nursing, said, "We identified 4,046 references that do not actually exist in peer-reviewed medical papers," in a study published in the Lancet on Apr. 9.
◇246 papers had three or more fake references
The Columbia team developed an artificial intelligence (AI) verification system to examine 125.6 million references cited in 2.5 million papers on PubMed Central from Jan. 1, 2023 to Feb. 18 this year. PubMed Central is a free databases of life science papers.
The AI verification focused on 97 million references with a Digital Object Identifier (DOI), a unique identifier assigned to digital content, or an ID assigned by the PubMed databases. The investigation found 4,046 false citations across 2,810 papers. The analysis showed 2,564 papers contained one or two fake references, and as many as 246 papers contained three or more fake references.
Cases of fake references in papers have spiked recently. In 2025, they were 12 times more frequent than in 2023. The researchers said, "The most rapid increase began in mid-2024, in step with the spread of AI writing tools." Kathryn Weber-Boer, a researcher at Digital Science in the United Kingdom, also suspected AI.
Speaking to Nature, Weber-Boer said, "We still do not know whether the fake references were fabricated by computers or by humans," but added that "the recent rise in such problems suggests the influence of Generative AI."
◇Could harm patient care
An increase in fake references is a serious problem. If the literature cited to support a paper's findings is fake, the paper's scientific basis is inevitably weak. A vicious cycle in which fake papers beget more fakes is unavoidable. In particular, because medical papers are directly tied to patient treatment, the consequences could be fatal.
Topaz said, "Because clinicians make treatment decisions based on clinical guidelines reported in papers, this directly affects patients." In other words, papers that cite fake studies can steer patients toward inappropriate care. Topaz added, "In one paper we reviewed, 18 of the 30 references were fake," noting, "some of those have already been cited by other papers, and they have appeared in review articles that inform clinical practice."
Even so, journals apparently took no action. The researchers said that, at the time of the investigation, 98.4% of the problematic papers had received no action from publishers. In a commentary published the same day, Howard Bauchner of Boston University School of Medicine and Frederick Rivara of the University of Washington School of Medicine said, "This study showed the need to improve research ethics."
Based on these findings, the Topaz team recommended that publishers verify references before accepting papers. They also suggested that companies providing indexing services supply additional information so users can assess the accuracy of references.
References
Lancet (2026), DOI: https://doi.org/10.1016/S0140-6736(26)00603-3