Researchers have developed an artificial intelligence tool that sifts through 2.6 million cancer studies published over the past 25 years and flagged more than 250,000 papers for closer scrutiny. The findings, published in The BMJ, come as scientific publishers face a surge of manuscripts produced by so-called paper mills-businesses that sell fabricated or manipulated research. The flagged papers are not confirmed frauds, but they share linguistic fingerprints with known paper mill products.
How the screening tool works
The Queensland University of Technology team trained the system using BERT, a language-processing model that recognises recurring writing patterns in previously identified paper mill publications. In testing, it correctly spotted known paper mill papers about 91% of the time. The researchers compare it to a spam filter: it doesn't decide whether an email is malicious, but it flags messages that deserve another look. Three scientific journals are already testing the technology as part of their editorial process. The approach reflects a broader shift toward using AI for Science & Research to automate the detection of problematic manuscripts.
A steady climb over two decades
The proportion of potentially problematic cancer studies rose from around 1% in the early 2000s to more than 16% by 2022. The trend appeared across thousands of journals, with molecular cancer biology and laboratory-based research showing some of the highest concentrations. Earlier this month, Nature reported that cancer papers suspected of originating from paper mills were attracting significantly more citations than legitimate studies. Researchers warned that questionable papers can spread through the scientific record as other scientists unknowingly cite and build upon them.
Why this matters for science and research professionals
Cancer studies influence everything from laboratory research and clinical trials to drug development and treatment guidelines. If unreliable research slips through, it can divert funding, mislead future studies and slow scientific progress. For professionals engaged in Research, the proliferation of questionable studies poses a direct threat to the reliability of the literature they depend on. The QUT team stress that being flagged by the AI is not evidence of misconduct-it is simply a signal that a paper deserves a closer look. As generative AI makes it easier to produce convincing scientific writing, tools that help editors focus their attention where it's needed most will become essential.
Your membership also unlocks:
