This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility: Months after a Columbia University School of Nursing–led audit published in the Lancet reported thousands of fabricated citations in medical papers, the same research team has found the same problem in the physical sciences, engineering and computing. About 1 in 435 science and engineering papers from 2026 cite studies that cannot be found online, up from 1 in 1,600 in 2023.
While relatively few papers were affected, fabricated references were identified across every discipline examined: 25.6 per 10,000 papers in physics, 18.4 in materials science, 11.1 in computer science and artificial intelligence, 10.3 in engineering and 4.3 in chemistry. The study, "Evaluation of possibly fabricated references in 500,000 science, engineering, and computing papers," is published in the Proceedings of the National Academy of Sciences (PNAS) and is part of a body of research made available this year by the team, examining this problem. A Lancet research letter in May, the first, identified 4,046 fabricated citations in 2,810 of 2.5 million medical papers.
A July paper in the Journal of Internal Medicine reported that fabricated references were about twice as common in medRxiv preprints as in peer-reviewed articles (37.2 versus 16.4 per 10,000). And a September arXiv preprint from the team, not yet peer reviewed, found 26 large language models fabricated 55% of the medical references they were asked to supply. "After our Lancet audit, the question we heard most often was whether this was a problem particular to medicine.
It is not," says Maxim Topaz, Ph.D., associate professor at Columbia Nursing and Columbia University Data Science Institute, who led the PNAS study. "A reference is a promise to the reader that a source exists. In a growing number of papers across the sciences, that promise is not being kept." The researchers analyzed 500,000 papers in chemistry, physics, materials science, engineering and computer science published from January 2023 through March 2026, drawn from OpenAlex, a large database of scholarly works.
An automated verification tool checked the 21.6 million references that carried a digital object identifier (DOI). It classified 1,135 references across 619 papers as fabricated when the cited title could not be found in online databases. When two independent experts reviewed a random sample of 400 flagged references, they confirmed 85.8% were fabricated.
The share of papers affected rose from 1 in 1,600 papers (6.3 per 10,000) in 2023 to 1 in 650 (15.4 per 10,000) in 2025 and 1 in 435 (23.0 per 10,000) in the first quarter of 2026. Most affected papers contained one or two fabricated references. But 27 papers contained five or more, and those 27 accounted for more than a third of all fabricated references identified.
Extract — continue reading at the source.
