Compatibility of emerging AI regulation with GATS and TBT: the EU Artificial Intelligence Act

Abstract Governments have recently started to design policies that are specific to artificial intelligence (AI), which is projected to become the dominant technology in the decades to come. AI is increasingly permeating all aspects of the digital economy, including trade in goods and services, giving rise to concerns whether emerging AI-specific regulation may run afoul of international economic law (IEL). However, studies on the law of the World Trade Organization have yet to pay close attention to the European Union (EU) Artificial Intelligence Act (AI Act), the first binding regulation of its kind in the world. This paper seeks to address this gap in the literature. It examines the compatibility of emerging AI regulation with multilateral trade rules, using the EU AI Act as a case study. More specifically, it analyses to what extent the EU AI Act’s disciplines on prohibited AI systems are likely to violate the EU’s existing obligations and commitments enshrined in the Agreement on Technical Barriers to Trade and the General Agreement on Trade in Services. This paper demonstrates that there is potential for conflict between the EU regulation and these two multilateral trade agreements. It also suggests that, although emerging AI regulation can represent a challenge for the future of IEL, the latter can play a role in guiding and shaping AI-specific regulation moving forward.