AI flags 250,000 cancer studies as scientists warn of fake research surge

Artificial intelligence is now being used to tackle a growing problem in science itself.

Researchers have developed an AI 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, don't suggest those studies are fraudulent. What they do suggest is that many share the same linguistic fingerprints as papers linked to so-called "paper mills" — businesses that produce or sell scientific manuscripts, sometimes using fabricated or manipulated data.

That distinction matters.

The AI is designed to raise questions, not answer them. Every paper it flags still needs to be reviewed by experts before any conclusions can be reached.

The research team at Queensland University of Technology (QUT) trained the system using BERT, a language-processing model capable of recognising recurring writing patterns in previously identified paper mill publications. During testing, it correctly identified 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; it simply spots the ones that deserve another look. Three scientific journals are already testing the technology as part of their editorial process.

What surprised the researchers wasn't just the number of papers flagged. It was the direction of travel.

Their analysis suggests that the proportion of potentially problematic cancer studies has steadily climbed over the past two decades, 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.

The study lands as publishers face mounting pressure to protect the integrity of scientific literature.

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.

That has consequences beyond academia.

Cancer studies influence everything from laboratory research and clinical trials to drug development and treatment guidelines. If unreliable research slips through the system, it can divert funding, mislead future studies and slow scientific progress.

Publishers have been tightening safeguards for several years, introducing stricter peer review, image-forensics software and plagiarism detection. AI is now becoming another layer of defence.

The authors stress that the technology is not meant to replace human judgement. Instead, they see it as a tool that can help editors focus their attention where it's needed most.

And they offer one final note of caution: being flagged by the AI is not evidence of misconduct. It is simply a signal that a paper deserves a closer look—an increasingly important distinction as generative AI makes it easier than ever to produce convincing scientific writing.