Predictive tools for the risk of criminal recidivism based on artificial intelligence: Towards their compatibility with the privacy and defense of the accused in the oral trial

Abstract

Artificial intelligence applied to the judicial process, has had as one of its central elements the introduction of software predictive tools for the risk of criminal recidivism at various stages of the procedural iter. Since its incorporation, the main conflicts that dispute with a variety of fundamental rights, such as the fundamental right to privacy and the fundamental right of defense of the indicted have been shown. This paper, focus in the oral trail stage, provides an explanation of the main nodes of this complex relationship; A configuration diagram of its violation is portrayed, addressing the procedural impact, and based on the above, possible conditions are proposed to make its incorporation compatible, without resulting in a violation of the aforementioned rights.