Trust and Distrust of and towards Artificial Intelligence Systems

Policy makers, scientists, and the general public are increasingly confronted with thorny questions about the regulation of artificial intelligence (AI) systems. A key common thread concerns whether AI can be trusted and the factors that can make it more trustworthy in front of stakeholders and the general public. This is indeed crucial, as trust is fundamental for both democratic governance and the development of AI. This piece, however, takes the reverse perspective: it explores how and the conditions by which AI systems could be considered as placing trust or distrust in humans – and more specifically, regulators, corresponding indeed to particularly relevant actors in charge of overseeing AI systems, as well as the implications of these patterns of trust and distrust. This is a largely overlooked and yet extremely relevant question because trust is a relational property, and AI systems deserve to be regarded – at least in some respects – as a regulatory target that is capable of some form of agency, for instance, by adjusting its behaviour to signals received by those in charge of designing and enforcing regulation.