Lombroso 2.0: On AI and predictions of dangerousness in criminal justice

The purpose of this paper is to analyze the legal and ethical issues raised by the use of artificial intelligence (AI) technologies in predicting criminal behavior. In fact, ever since Cesare Lombroso’s L’uomo delinquente, scientists, on one side, and jurists, on the other, have been discussing the ‘criminal brain’. Violence risk assessment tools have been applied in criminal courts for more than sixty years, yet the discussion has found once again importance thanks to the development of new AI techniques (such as machine learning) and to their application in both the medical and the criminal justice area. The new frontier is represented by the enhancement of these instruments through the combination of AI and neuropredictions. This paper presents critical reflections on the benefits and drawbacks of applying these technologies to predictions of violence in criminal justice. The inquiry concludes with a number of open questions which are hoped will contribute to the ongoing debate and work as a primer for future investigations on the matter.