A new study from Northwestern University warns that organized scientific fraud is becoming increasingly common. Researchers say organized groups manipulate the academic publishing system, from fabricated data to author purchases and paid citations.
To investigate this issue, scientists combined extensive analysis of scientific publications with detailed case studies. Although fraud is often portrayed as the result of omissions on the part of individual researchers, a team of researchers at Northwestern University discovered something much more complex. Their findings revealed a global network of people and organizations working together to systematically exploit weaknesses in the publishing process.
The scale of the problem is staggering. Researchers say fraudulent research is now being published at a faster pace than legitimate scientific publications. The authors say the findings should alert the scientific community to tighten safety measures before public trust in science begins to erode.
This study Proceedings of the National Academy of Sciences.
“Science needs to better monitor itself to preserve its integrity,” said the study’s lead author, Luis A.N. Amaral of Northwestern. “If we don’t raise awareness of this problem, more and more egregious behavior will become the norm. At some point it will be too late and the scientific literature will be completely poisoned. Some people worry that by talking about this problem we are attacking science. But we strongly believe that we are protecting science from bad actors. We need to recognize the seriousness of this problem and take action.”
Amaral studies complex social systems and is the Erastus Otis Haven Professor of Engineering Science and Applied Mathematics at Northwestern University’s McCormick School of Engineering. Reece Richardson, a postdoctoral fellow in Amaral’s lab, is the study’s lead author.
Investigating scientific fraud networks
When the public hears about scientific fraud, the focus is often on individual incidents, including data falsification, plagiarism, and retraction of research. These cases usually involve a single researcher who attempts to advance his career by taking shortcuts in a highly competitive environment.
But Amaral and his colleagues discovered a much more extensive and largely hidden system. Their analysis revealed a large underground network operating largely hidden from public view.
“These networks are essentially criminal organizations, acting together to falsify the scientific process,” Amaral said. “There are millions of dollars involved in these processes.”
To understand how widespread this problem is, the team examined a large collection of scientific data. This included a record of retracted papers, editorial information, and examples of reproduced images. Much of the information was derived from major scientific databases such as Web of Science (WoS), Elsevier’s Scopus, the National Library of Medicine’s PubMed/MEDLINE, and OpenAlex, which includes data from Microsoft Academic Graph, Crossref, ORCID, Unpaywall, and other institutional repositories.
The researchers also collected a list of de-indexed journals. These are academic journals that have been removed by the database for not meeting quality or ethical standards. Additional sources included records of retracted studies from Retraction Watch, discussion comments from PubPeer, and article metadata such as editor name, submission date, and acceptance date from selected journals.
Business of Paper Mills and Counterfeit Research
After analyzing the data, researchers identified coordinated activity involving paper mills, brokers, and compromised journals. A paper mill functions like a production line for academic papers. They produce large quantities of papers and sell them to researchers who want to quickly increase their paper record.
These manuscripts often contain fabricated data, manipulated or plagiarized images, plagiarized text, and sometimes scientifically impossible claims.
“More and more scientists are getting caught up in paper mills,” Amaral said. “Not only can they buy papers, but they can also buy citations. That way they can look like reputable scientists even though they’ve done very little of their own research.”
“Paper mills operate in a variety of models,” Richardson added. “So we’ve only been able to scratch the surface of how they operate. But they sell basically anything they can use to cleanse their reputation. They often sell author slots for hundreds or even thousands of dollars. People may pay more money for a first author position and less money for a fourth author position. People can also pay to have their papers automatically accepted by journals through a sham peer review process.”
To discover additional papers created through these operations, Amaral’s group launched another project to automatically scan published materials science and engineering research. The system searches for authors who have incorrectly identified the equipment used in their experiments. The findings from that study have been accepted for publication in the journal pro swan.
Brokers, journal hijacking, and organized fraud
The research team found that fraudulent networks rely on several strategies to disseminate fake research.
- A group of researchers collaborate to publish papers in multiple journals, even though their research is fraudulent. The paper is later retracted when misconduct is revealed.
- Brokers act as intermediaries who arrange for the publication of fraudulent papers in compromised journals.
- Fraud often focuses on specific scientific fields that are susceptible to manipulation.
- Organized groups find ways to circumvent quality control measures such as journal indexing.
“Brokers connect different people behind the scenes,” Amaral said. “You need to find people to write the paper. You need to find people who will pay you to be authors. You need to find a journal to publish the entire paper. And you need an editor at that journal to accept it.”
In some cases, these groups avoid legitimate journals altogether and instead take over abandoned journals. When a legitimate publication ceases to operate, fraudsters can acquire the website or domain name and reinstate it as a vehicle for fraudulent publication.
“This happened with the magazine ‘HIV Nursing,'” Richardson said. “It used to be the journal of a British professional nursing organization, but it has since ceased publication and its online domain has expired. An organization bought the domain name and began publishing thousands of articles on subjects completely unrelated to nursing, all indexed in Scopus.”
protect the integrity of science
Amaral and Richardson say the scientific community needs a broad strategy to deal with the growing threat. This includes closer monitoring of editorial practices, powerful tools to detect fabricated research, a deeper understanding of the networks that enable fraud, and significant changes to the incentive systems that encourage scientific publishing.
The researchers also emphasize the urgency of addressing these issues before artificial intelligence (AI) becomes more deeply integrated into the scientific literature.
“If we’re not prepared to deal with the fraud that’s already happening, we certainly aren’t prepared to deal with the potential impact of generative AI on the scientific literature,” Richardson said. “We have no idea what will end up in the literature, what will be considered scientific fact, what will be used to train future AI models, and what will then be used to write further papers.”
Amaral said the project is personally disappointing but necessary.
“This study is probably the most depressing project I’ve ever been involved with,” Amaral said. “I’ve been excited about science since I was a child. It’s sad to see others commit fraud or mislead others. But if you believe that science is useful and important to humanity, you must fight for it.”
“The organizations that enable large-scale scientific fraud are large, resilient, and rapidly growing,” the study was supported by the National Science Foundation and the National Institutes of Health.
