AI flags more than 250,000 suspicious cancer research papers

A machine learning system analyzed 2.6 million cancer research papers and identified over 250,000 as suspicious. The system detected writing patterns similar to those produced by fraudulent "paper mills."

The machine learning system compared linguistic features to known paper-mill outputs. The specific patterns flagged were not detailed in the report, but the tool appears to focus on stylistic anomalies.

Paper mills are entities that sell fabricated research papers to authors seeking publication. The use of AI to detect such fraud is an emerging and expanding field. Researchers hope these tools can help maintain the integrity of scientific literature and plan to refine the system for broader use.