The advent of intelligent systems in healthcare produces significant benefits, including resource efficiency and "personalization" of the therapeutic experience. However, the overwhelming spread of artificial intelligence (AI) in diagnostics will expose businesses and individuals to a diverse range of operational and legal risks. The following paper focuses on the phenomenon of "data poisoning," a cyberattack likened to the beating of butterfly wings capable of generating high-impact scenarios. 

Policy Brief n. 16/2022, Intelligenza artificiale e diagnostica medica. Tra nuove sfide ed “effetto farfalla”, Giovanni Alliegro, Fosco Gentili, Giuseppe Modugno, Alessandro Rondinone