Including a Social Perspective In AI Ethics: The Contribution of a Dialogue Between American Pragmatism and Critical Theory

Abstract

            Throughout the history of moral philosophy, the theoretical postures have been privileged. Modern ethics is no exception and is indeed characterized by the predominance of voluntarist and universalist frameworks, which are primarily concerned with the actions of the moral agent, with no real regard for the conditions of possibility necessary for the effective realization of moral actions. Recent developments in applied ethics have shown that an integral application of classical ethical frameworks does not adequately address the new moral dilemmas emerging from our different spheres of activity. Artificial intelligence (AI) ethics once again demonstrates the inadequacy of traditional ethical frameworks to deal with the many ethical issues related to the pervasiveness of AI systems. Indeed, the dominant theories in ethics fail to take account of the shared responsibility that characterizes the moral obligations we have towards AI systems. The particularity of pragmatist ethics is that it aims at a practical intervention without however renouncing the conceptual clarifications necessary for such an intervention. We will demonstrate how the characteristics of pragmatist ethics avoids certain pitfalls in AI ethics and provides a conceptual framework particularly well suited to address the ethical issues related to the increasing use of AI systems in our societies.