النظام القانوني للمسؤولية التقصيرية عن أخطاء الذكاء الاصطناعي: دراسة استشرافية

Abstract

This study highlights the legal complexities associated with tort liability arising from artificial intelligence (AI) system errors, aiming to evaluate the capacity of traditional civil liability rules to accommodate current technological advancements. To achieve this objective, the study adopts an analytical and prospective approach to test the efficacy of traditional theories such as vicarious liability, the custody of things, and product liability highlighting their structural shortcomings in dealing with technologies characterized by autonomy and self-learning in the absence of an independent legal personality for these digital entities. Consequently, the research concludes that it is practically impossible to adapt traditional legal texts to contain the algorithmic leap without undermining their established concepts. This necessitates a transition towards modern approaches, foremost among them the Human Agent theory innovated by the European legislator as a novel accountability framework that exempts the injured party from the burden of proving technical fault. The study concludes with fundamental recommendations emphasizing the imperative to move beyond slow, traditional litigation mechanisms and to establish automatic and flexible redress mechanisms. This requires legislative intervention to impose a mandatory insurance system on A developers and users, and the creation of reserve compensation funds to effectively protect victims' rights, thereby striking the desired balance between repairing harm and fostering technological innovation.