Artificial intelligence is a subject from which the civil law scholar draws more than one reason for discouragement, not only because of the intrinsic intertwining with knowledge and disciplines of a technical nature, to which the civil jurist is not usually accustomed, but above all because of the speed with which artificial intelligence has entered the spaces of Law, exceeding its capacity to react and leaving numerous questions unanswered. Among these, perhaps more than others, is the question of liability for damages related to the use of artificial intelligence: even the algorithm can make mistakes. In this context, unprecedented questions of liability arise in the event of harmful consequences for third parties resulting from automated decision-making processes using algorithms. In this regard, if the liability of the producer and the user is taken into account, a possible causal contribution of the algorithm's programmer must also be considered: in order to ensure maximum protection for the injured party, liability models based on fault do not seem to have a place, but the application of the stricter strict liability regime seems more rational. The study stops at verifying the possible applicable rules.
L’intelligenza artificiale è tema dal quale lo studioso del diritto civile ricava più di una ragione di sconforto, non solo per la connaturale commistione con cognizioni e discipline di carattere tecnico, alle quali il civilista è di solito poco avvezzo, ma soprattutto per la velocità con cui l’intelligenza artificiale ha fatto ingresso negli spazi del diritto, superandone la capacità di reazione e lasciando senza risposta numerose questioni. Tra queste, forse più di altre, emerge quella della responsabilità per i danni legati all’utilizzo dell’intelligenza artificiale: anche l’algoritmo può sbagliare. In tale quadro, si prospettano inedite questioni di responsabilità per il caso che dai processi decisionali automatizzati tramite algoritmi derivino conseguenze dannose a carico di terzi. In proposito, se viene in considerazione la responsabilità del produttore e quella dell’utilizzatore, deve anche prendersi in considerazione un possibile contributo causale del programmatore dell’algoritmo: al fine di garantire al danneggiato massima tutela non sembra esservi spazio per modelli di responsabilità fondati sulla colpa, ma pare maggiormente razionale l’applicazione del più rigoroso regime della responsabilità oggettiva. Lo studio si sofferma sulla verifica delle possibile regole applicabili..
La responsabilidad por mal funcionamento de algoritmo en la legislación de la UE
elsa anna stefania bivona
2025-01-01
Abstract
Artificial intelligence is a subject from which the civil law scholar draws more than one reason for discouragement, not only because of the intrinsic intertwining with knowledge and disciplines of a technical nature, to which the civil jurist is not usually accustomed, but above all because of the speed with which artificial intelligence has entered the spaces of Law, exceeding its capacity to react and leaving numerous questions unanswered. Among these, perhaps more than others, is the question of liability for damages related to the use of artificial intelligence: even the algorithm can make mistakes. In this context, unprecedented questions of liability arise in the event of harmful consequences for third parties resulting from automated decision-making processes using algorithms. In this regard, if the liability of the producer and the user is taken into account, a possible causal contribution of the algorithm's programmer must also be considered: in order to ensure maximum protection for the injured party, liability models based on fault do not seem to have a place, but the application of the stricter strict liability regime seems more rational. The study stops at verifying the possible applicable rules.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.