REGULATING AI ACROSS BORDERS: THE CASE FOR A BINDING INTERNATIONAL FRAMEWORK

Abstract

This article argues that fragmented governance approaches to artificial intelligence are insufficient and that a binding international treaty is urgently required. Unlike existing soft-law initiatives, the proposed framework is rooted in enforceability, adaptability, and multilateral legitimacy. Drawing from lessons in arms control, climate accords, and data protection, the paper advances original design pathways, including tiered obligations, sector-specific annexes, and the creation of a Global AI Observatory. The analysis demonstrates that without such a treaty, regulatory arbitrage, inequities between the Global North and South, and risks to fundamental rights will intensify. The contribution lies not only in diagnosing governance failures but in proposing a practical architecture for binding global regulation. In doing so, the article extends scholarly debates beyond descriptive accounts toward normative institutional design. The findings are relevant to policymakers, international lawyers, and scholars concerned with embedding human rights and equity in the governance of emerging technologies.