Ethics implications of the use of artificial intelligence in violence risk assessment

Artificial intelligence is rapidly transforming the landscape of medicine. Specifically, algorithms powered by deep learning are already gaining increasingly wide adoption in fields such as radiology, pathology, and preventive medicine. Forensic psychiatry is a complex and intricate specialty that seeks to balance the disparate approaches of psychiatric science, which strives to explain human behavior deterministically, and the law, which emphasizes free choice and moral responsibility. This balancing, a central task of the forensic psychiatrist, is necessarily fraught with ambiguity. Such a complex task may intuitively seem impenetrable to artificial intelligence. This article first aims to challenge this assumption and then seeks to address the unique concerns posed by the adoption of artificial intelligence in violence risk assessment and prediction. The relevant ethics concerns are analyzed within the framework of traditional bioethics principles. Finally, recommendations for practitioners, ethicists, and others are offered as a starting point for further discussion.