This recent PNAS study shows that large-scale scientific fraud is driven by coordinated networks of editors, authors, brokers, and “paper mills” that mass-produce and strategically place fabricated research. Case studies from journals like PLOS ONE, Hindawi, and IEEE conferences reveal that certain editors repeatedly handle fraudulent submissions, often from the same authors, and that fraudulent papers cluster by time and publisher—evidence of organized production. One group, ARDA, was found to “journal hop” when venues were deindexed, ensuring continued publication opportunities, sometimes in unrelated fields.

Fraud is especially prevalent in certain research areas, such as microRNAs, circular RNAs, and long noncoding RNAs, where retraction rates are far higher than in related subfields. Suspected paper mill output is doubling every 1.5 years—much faster than legitimate science—and already exceeds the reach of current integrity measures. The authors conclude that without robust, conflict-free systems for detection and enforcement, large portions of the scientific record risk being permanently compromised.