Psychiatric-forensic evaluations and artificial intelligence: new possible scenario

Abstract

Cognitive biases are defined as mental processes which can lead to the elaboration of misjudgements. Biases can influence thoughts, opinions, behaviours, and they are inevitably involved in psychiatricforensic evaluations. The delicate role that the mental health expert plays in the psy-chiatric examinations makes these mistakes highly relevant. Make sure that such a complex decision-making process is set up with care and attention is fundamental to guarantee the best protection of the individual rights, but also to avoid a mismanagement of government money. In order to avoid the possible use of cognitive biases in psychiatric-forensic assessments, it has been suggested to increase the standardization of evaluation procedures, but scientific literature shows that experts working in this area are often reluctant to question their own work. Recently, the application of Artificial Intelligence (AI) to the forensic field has opened new possibilities, but, on the other hand, it has generated new questions. AI seems to be beneficial for those decision-ma-king processes where greater standardization is required, but attempts to use AI tools in the field of forensic psychiatry have highlighted some critical issues. In this article, after discussing the problems that characterize both human and computational decision-making, we will propose possible solutions.