La “scatola nera” della decisione giudiziaria: tra giudizio umano e giudizio algoritmico

Abstract

The increasing role of artificial intelligence has confronted judicial systems with the perils of algorithmic decisions. Facing the prospect of “robot-judges”, the right to a fair trial still relies on human decisions. However, the mechanisms of judicial decision-making need to be questioned. In fact, it can be said that if algorithms operate as “black boxes”, so does the mind of human judges. According to the most extreme version of legal realism, «law is what the judge had for breakfast»: opaque reasoning and bias not only affect algorithmic decisions, but human decisions, too.