La giustizia predittiva : riflessioni sparse

The use of artificial intelligence also in the judicial field offers undoubted advantages, as it can help speed up proceedings and the ensuing judgement, especially in cases where purely technical operations are required. However, its widespread use, especially in the ‘predictive’ function of judicial decisions, raises many concerns. In fact, the risk could be both that of hindering the necessary evolution of the law (crystallised in the repetition of already elaborated models) and of distorting the interpreter’s indispensable activity also linked to the peculiarities of concrete cases. It must also be borne in mind that artificial intelligence, operating by means of algorithms instructed through data input, offers solutions that are only apparently objective and neutral, since they may actually reflect the choices, interests, and ideologies of the programmer who devised the code or set up the system, or of those who selected the elements to be inserted. Hence the need to regulate the phenomenon (as called for by much doctrine and being implemented by the European legislator).