Soft Law Governance in the Field of AI: A European Perspective

Recent years have seen the emergence of a multitude of soft law initiatives aimed at governing artificial intelligence (AI) technologies and their perceived societal risks. This chapter engages a Foucauldian governmentality perspective to provide an analysis of the regulatory logics at play in particularly European attempts to (1) conceptualize AI as an object of governance, and (2) propose concrete mechanisms for its effective governance. Acknowledging the evolving state of the field, we illustrate and problematize increasingly institutionalized approaches to AI governance, including AI risk management, principle-based governance, and AI certification and labelling. While the Foucauldian approach is helpful in constructing an initial outline of this emergent governmental field, the chapter also highlights the significant ambiguities and uncertainties surrounding AI governance. We therefore think it important for regulators, scholars, and other interested or affected parties to prioritize the question of AI governance in years to come.