The Institutional Design of International Organizations in the Global Governance of AI

The promises and pitfalls of Artificial Intelligence (AI) have unleashed a “race to regulate AI.” As AI’s development and deployment generate transnational externalities that require international cooperation, international organizations (IOs) play a pivotal role in facilitating these cooperation efforts. While no international organization has emerged as the central front-runner in leading the regulatory efforts, several international or regional institutions have positioned themselves as key institutions in the emerging global AI governance landscape. Yet these IOs vary considerably in terms of their institutional characteristics and IO design is likely a key determinant in whether and how IOs shape the emerging global governance of AI. The article makes several contributions to the literatures on IO design and the global governance of AI. Theoretically, while earlier research on institutional design has shown how state actors and bureaucratic elites shape IO design, we know little about their preferences vis-à-vis AI governance, an entirely new issue area. Further, this article theorizes the impact of firms on IO design, which leverage considerable influence due to their superior technical expertise. To this end, the article introduces the concept of an elite-firm-citizen gap in global AI governance. Empirically, it extends a new dataset on the global governance of AI to explore IO design dimensions and utilizes quantitative text analysis methods to uncover the preferences of elites, firms, and citizens on the institutional features of AI governance institutions.