Brevi riflessioni sull’utilizzo dell’intelligenza artificiale da parte del giurista

The paper analyzes the implications of using AI systems in the exercise of administrative action, highlighting the problematic issues it raises and outlining possible interpretative solutions in the light of national and European regulatory frameworks. It examines the basis and the limits of employing AI systems within administrative procedures, both in cases of discretionary activity and of technical-discretionary activity (the so-called “algorithmic legality”), as well as the need to ensure, in order to protect the interests involved, an indispensable (final) review by a human official. The study aims, in particular, to show that the activity carried out by AI cannot be considered neutral and, consequently, to emphasize the importance of providing public oversight over the data input process, as well as over the resulting general categorization activity that information systems (inevitably) perform in order to reach an automated decision.