AI tool flags over 250,000 cancer papers as possible paper-mill fakes
A new AI tool built by researchers at Queensland University of Technology scanned 2.6 million cancer papers and flagged more than 250,000 that share the linguistic fingerprints of studies produced by paper mills. The suspect share rose from about 1 percent in the early 2000s to o
Cancer patients and doctors rely on published research to choose treatments. A new analysis published in The BMJ, led by Professor Adrian Barnett of Queensland University of Technology, used an AI tool to scan 2.6 million cancer papers published between 1999 and 2024. It flagged more than 250,000 that share the linguistic fingerprints of studies produced by so-called paper mills, businesses that sell manuscripts, sometimes with fabricated or manipulated data. The tool does not prove any single paper is fake; it identifies writing patterns matching known problem papers, and each flagged study still needs expert review. The suspect share climbed from around 1 percent in the early 2000s to more than 16 percent by 2022. Reviews, treatment guidelines, and future experiments build on this published record, so contamination at this scale can quietly influence both research and clinical decisions. Sources: The BMJ; QUT; ScienceDaily.