The Paris Journal on AI & Digital Ethics

Abstract

This paper is dedicated to introduce the fourth thematic of the Paris Journal on AI & Digital Ethics: Rebuilding social cohesion in AI-powered societies. We situate the full papers of the track in a broader context: what do they tell us about trust in the era of AI? The Introduction 1 opens with a paradox concerning trust, between the presence of the word and the absence of the plain phenomenon. If trust is called upon, it is that the notion behind needs to be restored. We then distinguish between two types of trust, depending on the considered scale. At the political or social level, AI tools play a classical dual role to promote participatory governance or increase the opacity of decisions. At the individual level, AI bots and recommendations can be treated as trusted third parties, when effectively interacting with people. As applying interpersonal trusting effects to AI models is not obvious, we discuss briefly in Section 2 how moral, affective and cognitive mechanisms might be mirrored from human to machine trustees. Through our lens on trust, we finally propose a view on the papers of the track. From philosophical theories to empirical studies, from social to cognitive frameworks, we argue in Section 3 that they complete each other to picture the new landscape of social trust mediated by AI, and individual trust in AI devices.