From judgment to the algorithm: the risk of a justice system without humanity

Abstract

In this work, I examine the increasing use of artificial intelligence in criminal justice, with particular reference to predictive policing and algorithmic decision-making systems. The main claim I set out to defend is that the integration of algorithmic systems into judicial processes should not be understood as a merely technical development, but rather as a transformation affecting the very meaning of legal judgment. This means that the adoption of those technologies threatens to marginalize the human dimension of adjudication and to replace interpretative and relational practices with forms of statistical calculation and predictive reasoning. On this basis, I reconstruct some of the main theoretical conceptions of legal adjudication, in order to highlight the role of discretion and interpretation in judicial decision-making. I conclude by arguing that the claim of a fully predictable and computable justice may well result in a dehumanized model of adjudication.