Artificial intelligence and civil liability: regulatory perspective in the light of the EU Artificial Intelligence Act

Abstract

Artificial Intelligence systems are used in several applications. Recently they have been applied in the medical and legal fields, including the criminal one. Coherently with the diffusion of these systems, the need for their regulation has emerged, especially in the civil liability field. In this essay, we will provide a perspective of the future European regulation concerning the civil liability of artificial intelligence systems, starting from the genesis of the European Union legal framework on the subject. We will consider the issues of the AI definition and the classifications made between high-risk AI and other Artificial Intelligence types. The resulting framework is then compared to the Regulation proposal contained in the European Commission’s Artificial Intelligence Act, to assess correlations and differences, and define possible developments of the matter.