An AI-assisted audit by an international, multidisciplinary research team reveals that nearly 3,000 peer-reviewed medical articles contain fabricated references that do not exist in any scientific database. The findings, published as a research letter in The Lancet, point to a worrying shift driven using artificial intelligence in academic writing.
The University of Eastern Finland was part of the research team, which developed an AI-based automated verification system. The system scanned 2.5 million articles from PubMed Central's Open Access collection published between 1 January 2023 and 18 February 2026. Among 97.1 million verified references, the team identified 4,046 fabricated citations across 2,810 articles. The number of fabricated references has grown more than 12-fold since 2023, with the sharpest rise beginning in mid-2024, aligning with the mainstream adoption of AI-based writing tools.
“The findings are alarming from a patient safety perspective. Clinical guidelines and systematic reviews rely on the assumption that the studies they cite are real and trustworthy. If fabricated references slip into the evidence base, errors can travel into patient care unnoticed. We now need rapid action from both publishers and the research community,” says Associate Professor Laura-Maria Peltonen of the University of Eastern Finland.
The research team emphasises that health professionals and clinical guideline developers have no way of knowing whether the evidence they rely on is real. In one article containing fabricated references that the team examined, 18 out of 30 citations were fake. Some of these citations have already been cited in other articles and appear in systematic reviews that inform clinical care.
Based on their findings, the researchers recommend that publishers verify the references in every manuscript. They also recommend that indexing services add metadata to records, allowing users to assess the validity of references.
The team further urges major research integrity databases to establish a dedicated category for fabricated references, to enable systematic tracking and accountability. The researchers also call on publishers to retroactively screen already-published articles and to issue corrections or retractions when fabricated references compromise a paper's conclusions. Notably, at the time of the audit, no publisher action had been taken in 98.4% of the articles in which fabricated references were identified.
In an accompanying commentary published alongside the study, Professor Howard Bauchner (Boston University Chobanian and Avedisian School of Medicine) and Professor Frederick P. Rivara (University of Washington) state that the study sheds light on a deeply concerning phenomenon and underscores the need to maintain and strengthen research integrity:
“Given that public trust in science appears to be waning in countries around the world, renewed efforts are needed to enhance research integrity. Authors must take responsibility and be held accountable for the entire content of a manuscript, including the references.”
The project was led by Maxim Topaz of Columbia University. The research team included Dr. Nir Roguin (Tel Aviv Sourasky Medical Center and Faculty of Health Sciences, Ben-Gurion University of the Negev), Dr. Pallavi Gupta and Dr. Zhihong Zhang (Columbia University School of Nursing), and Associate Professor Laura-Maria Peltonen (Department of Health and Social Management, University of Eastern Finland; Wellbeing Services County of North Savo; Faculty of Medicine, University of Turku; and the Wellbeing Services County of Southwest Finland).
For more information:
Associate Professor Laura-Maria Peltonen, Tel. +358 50 590 3413, laura-maria.peltonen(at)uef.fi
Research letter:
Available online here.