In search of a global governance mechanism for Artificial Intelligence (AI): a collective action perspective

Abstract Proposals for governing Artificial Intelligence (AI) developments at a global level often neglect an important element—the divergent approaches of the United States, China, and the European Union to AI governance. However, the non-cooperation of these three key actors in recent AI developments poses barriers to global AI governance and risks to humanity. This situation constitutes a collective action problem. How, then, can a collective action perspective inform global AI governance efforts? Applying insights from common-pool resources (CPRs) and large-scale collective action studies, this study proffers considerations for designing institutions for global AI governance, factoring in the non-cooperation of the three key players. Findings suggest that a polycentric multilevel arrangement of AI governance mechanisms will be more effective than a single centralized global AI governance mechanism. Enforcement of AI governance rules is crucial for implementation success and should be supported by monitoring and sanctioning mechanisms, and conflict-resolution and information-provision mechanisms, with roles for international organizations.