ARTIFICIAL INTELLIGENCE AND HUMAN AUTONOMY

The emergence of artificial intelligence (AI) is giving rise to very prolific debates in applied ethics and arousing great social interest. However, they suffer from conceptual ambiguity and a lack of systematic development of the issues involved. Thus, the studies contributed to these debates are often superficial and limited to either indicating that AI systems will lead to desirable or undesirable consequences, or to deriving ethical conclusions from the mere analysis of particular AI systems.

The AutAI research project aims to overcome these shortcomings through an in-depth study of autonomy, one of the central concepts of ethics in general, and AI ethics in particular. We believe that this study will make it possible to break the impasse to which the different theories about it have led. Thus, from an eclectic conception of autonomy that brings together the virtues of different theories, we will try to determine different criteria for identifying the possible risks and benefits of AI for autonomy. These criteria will be applied to the study of three AI technologies in particular: recommender systems, social robots and virtual reality.

As a result of this preliminary work, an attempt will be made to formulate a proposal for ethical principles that will serve as a basis for a design and use of such technologies that are respectful and facilitative of human autonomy. Finally, the technical, socio-political and educational aspects of implementing the normative proposal reached will be considered.