Steering the governance of artificial intelligence: national strategies in perspective

Abstract

As more and more governments release national strategies on artificial intelligence (AI), their priorities and modes of governance become more clear. This study proposes the first comprehensive analysis of national approaches to AI from a hybrid governance perspective, reflecting on the dominant regulatory discourses and the (re)definition of the public-private ordering in the making. It analyses national strategies released between 2017 and 2019, uncovering the plural institutional logics at play and the public-private interaction in the design of AI governance, from the drafting stage to the creation of new oversight institutions. Using qualitative content analysis, the strategies of a dozen countries (as diverse as Canada and China) are explored to determine how a hybrid configuration is set in place. The findings show a predominance of ethics-oriented rather than rule-based systems and a strong preference for functional indetermination as deliberate properties of hybrid AI governance.