Every 10,000 physics papers there are 25 fake citations, finds new study

The findings were published in the PNAS by researchers at the Columbia University School of Nursing.

The findings were published in the PNAS by researchers at the Columbia University School of Nursing | Representative Image

New Delhi: Fabricated references are no longer just a problem in medical research and can be found across studies in the field of physics, engineering, chemistry, computer science, and material science.

The findings came out of a new audit of 500,000 scientific papers. The study was published in the Proceedings of the National Academy of Sciences (PNAS) and led by researchers at the Columbia University School of Nursing. Earlier this year, the team had published another audit reporting thousands of fabricated citations in medical papers.

“After our Lancet audit, the question we heard most often was whether this was a problem particular to medicine. It is not. 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,” lead author Maxim Topaz, associate professor at Columbia University, said in a statement.

For their study, researchers examined 500,000 papers published between January 2023 and March 2026 using OpenAlex, a large database of academic publications. Using an automated tool, they cross-checked nearly 21.6 million references.

The team found 1,135 references across 619 papers whose titles could not be found in any online database. When independent experts reviewed a random sample of those 400 flagged references, they too found that about 85 per cent of them were fake.

Physics at the top

Among the different fields, physics had the highest rate of fabricated references, at 25.6 per 10,000 papers. Material sciences came next at 18.4. Computer science and AI had 11.1 fake citations per 10,000 papers, and chemistry had only 4.3.

The authors estimate that nearly 8,500 papers published across these fields during the study period contain at least one fabricated reference. At the time of the study, only two of the 619 papers with fake citations had been retracted.

The new study follows an earlier audit published in May by the same team in The Lancet. At the time, the team had found 4,046 fabricated citations in 2,810 of 2.5 million medical papers.

“A paper with one unverifiable reference and a paper with five or more are not the same thing, and editors should not treat them the same way. The encouraging part is that this is one of the few integrity problems that can be checked automatically, before a paper is ever published,” Topaz said.