Piercing the Westphalian Veil: Assessing the Liability for AI-Driven Harms

Abstract

Provide As autonomous AI systems outpaced traditional Westphalian legal frameworks, a global Enforcement gap has emerged, leaving non-state-driven harm largely unchecked by international law. This article explores the imperative transition from ethical soft law to binding law accountability. The primary objective is to propose a Binding International AI Liability Protocol (IALP) under the 2025 UN Global Digital Compact, utilising the technical architecture of the EU AI Act and the Product Liability Directive (2024/2853). Employing a doctrinal research methodology, the study analyses the EU’s strict liability standards and the presumption of defectiveness as normative archetypes for global accountability. This research is significant as it provides a blueprint for bridging the digital-liability divide, ensuring victims in the Global South are protected from cross-border algorithmic harm. It concludes that by multilateralizing the EU’s liability regime through a UN-anchored agency, the international community can establish a universal social contract that holds multinational corporations directly accountable. Without this binding transition, global AI governance will remain fragmented, leaving human rights vulnerable to the functional sovereignty of tech giants.