Modern Artificial Intelligence (AI) poses new challenges for tort law. The ability of AI systems to learn and evolve from their initial programming is capable of generating significant regulatory gaps that law is called upon to address. On the one hand, there is no ex ante safety discipline related to the production phase of smart products; on the other hand, the characteristics of modern AI systems have the potential to challenge the tightness of existing national liability rules. The aim of the present research is to investigate the aforementioned protection gaps and the regulatory tools that the institutions of the European Union are preparing to fill them, in a perspective that necessarily has to consider the transition period that the study of the subject is going through, as well as the way in which the different levels of regulation are able to interact with each other.

The thesis examines the technological innovations related to artificial intelligence from the point of view of fundamental rights. The first part gives a general introduction on the technical characteristics of artificial intelligence, its historical development and the ethical, philosophical and legal debate about it. The second part analyses the impact of AI on the rights recognised by the Italian Constitution and the other main legal traditions based on the rule of law, focusing in particular on the protection of personal identity, the right to non-discrimination and freedom of speech. The third part explores the hypothesis of the recognition of three new rights due to the technological revolution: the right to artificial intelligence disclosure, the right to an explanation of outputs and the right to a human oversight of technology. These new legal positions are analysed in three hypothetical contexts of application: the public administration, the justice sector and the healthcare industry.