Paper mills pose a systemic threat to research integrity because they contaminate evidence, affect citations, and can influence clinical decision-making. These are companies that create manuscripts resembling legitimate scientific articles with fabricated, manipulated or duplicated data and images. These manuscripts are sold to authors and submitted to journals as actual papers. The mainstream media and leading scientific journals are increasingly reporting the scale of the problem and its implications for trust in science. Uncovering paper articles is challenging because they are dispersed throughout the journals and examined individually. Post-publication evaluation depends on a small number of specialized detectives. Some journals lack incentives to acknowledge big issues. The article deals with the use of AI to detect these papers.