Nearly 10% of cancer research papers published between 1999 and 2024 show signs of fabrication by “paper mills” selling manuscripts at industrial scale, according to a study published in BMJ and highlighted by researchers at the Kaiser Family Foundation (KFF) this week.
The study, which utilized a machine learning model to identify potentially fraudulent papers, found that the proportion of flagged papers rose exponentially over time, increasing from approximately 1% in the early 2000s to over 15% of annual cancer research output by the 2020s. The share of flagged papers appearing in high-impact journals also increased, exceeding 10% in recent years.
The findings come as lawmakers express increasing concern over research integrity and potential foreign influence in federally funded research. House Republicans sent oversight letters to five federal agencies earlier this month, demanding information on safeguards to prevent falsified studies from impacting grant allocations. The letters specifically cited concerns about paper mills with ties to the Chinese Communist Party, alleging that pressures on Chinese researchers may be driving demand for fabricated research. Publishers have already been forced to retract thousands of papers linked to these activities, and some have shut down journal subsidiaries after discovering widespread fraud.
Researchers are responding to the growing threat by turning to AI-powered detection methods. The BMJ study demonstrates the potential of machine learning to identify papers with textual characteristics similar to those produced by known paper mills. However, the KFF report notes that the problem may be intensified by the increasing sophistication of generative AI, which could be used to automate the creation of fraudulent papers.
Alongside concerns about fabricated research, officials are also grappling with the spread of misinformation regarding vaccine incentives. Texas Attorney General Ken Paxton recently announced an investigation into alleged “unlawful financial incentives” related to childhood vaccine recommendations, echoing claims that doctors are paid based on the number of vaccines administered. However, analyses show that pediatricians typically break even or lose money when administering vaccines, particularly to uninsured or Medicaid patients. Despite these claims, a recent KFF/Washington Post survey found that pediatricians remain the most trusted source of vaccine information among parents.
Further complicating the landscape, Health and Human Services Secretary Robert F. Kennedy Jr. Claimed earlier this month that the ketogenic diet could “cure” schizophrenia, a statement challenged by the Harvard researcher whose work Kennedy cited. The researcher clarified that they have not claimed a “cure” and do not advise patients to try the diet without medical supervision. Kennedy’s comments sparked a surge in online discussion about the ketogenic diet and mental health, reflecting broader uncertainty about medical evidence.
A separate study published in Nature Aging found that older adults are disproportionately exposed to low-credibility health information online, even though they consume less health content overall. This disparity highlights the need for targeted public health communication strategies to address age-related vulnerabilities to misinformation.
