The regulation of artificial intelligence for Europe: a normative and political project

Fabrice GUTNIK, AFNOR Consulting, Associate Professor and Researcher, Jules Vernes University of Picardy CURAPP UMR 7319, France

The European Commission envisions this decade as Europe's digital decade. Even more, the Commission's ambition is "now (to) strengthen its digital sovereignty and set standards instead of following those of others, with a clear focus on data, technology and infrastructure" . Its AI policy is part of the development of the Digital Single Market (DSM) and is embodied in its proposed 2021 Artificial Intelligence Regulation. According to some research, this text "envisages the regulation of AI through all activities whose economic processes are oriented towards the provision of goods and services related to AI (industrial activities): R&D activities in the form of patent applications for technological developments related to AI; and, more marginally (...) academic research activities related to AI" . However, while this draft regulation may have attracted some favorable comments, on the one hand, it has also been criticized for the lack of obligations to respect fundamental rights in a joint opinion of the European Committee and the European Data Protection Supervisor, published on June 18, 2021, or for its nuanced criticism by the CNIL in its opinion published on July 8, 2021 ; and, on the other hand, gave rise to new recommendations from the Commission Nationale Consultative des Droits de l'Homme (CNCDH), which unanimously adopted on April 7, 2022 an opinion entitled "Artificial Intelligence and Human Rights: For the development of an ambitious legal framework"". The purpose of this communication is therefore to report and discuss the ethical debates that the publication of the European AI regulation raises.