Global Standard-Setting for Artificial Intelligence: Para-regulating International Law for AI?

Abstract

Abstract Although there are several ongoing attempts at the international level to develop rules applicable to artificial intelligence (‘AI’) across countries, only the joint work of the International Organization for Standardization (‘ISO’) and the International Electrotechnical Commission (‘IEC’) has been successful at delivering a text with global character: the foundational standard, ‘ISO/IEC 22989:2022(E)—artificial intelligence concepts and terminology’, published in July 2022. This article explores the meaning and implications of this new standard from a regulatory point of view. The argument advanced is that despite its voluntary nature, this standard sets a pathway for future regulation of AI internationally, making it a paramount example of para-regulation. Para-regulation is understood here as the norms that do not place themselves within the legal space but delineate boundaries for posterior developments within that legal space. The author adopts a critical view of this para-regulatory role, arguing that the institutional arrangements in which these standards are created, in terms of representation and decision-making procedures, are far from ideal. This is particularly the case given the possibility of regulatory confrontation between key international players.