The Global Reach of the European Union’s Artificial Intelligence Act: An Empirical Process-Tracing Case Study of OpenAI and the Brussels Effect

As firms start implementing the first obligations under the European Union’s Artificial Intelligence Act, scholars predict potential a potential Brussels Effect, where companies voluntarily align with EU regulations beyond EU borders. This study investigates the reasons for external policy alignment with the AI Act by answering the research question: Why do non-EU actors align with the European Union’s Artificial Intelligence Act outside the EU? Using OpenAI as a case study, the research applies a process-tracing method to assess whether the five drivers of the Brussels Effect motivated OpenAI’s alignment with the AI Act. Primary and secondary document analysis was conducted, in combination with semi-structured interviews with five policy experts, allowing for the analysis of empirical data. A causal mechanism was constructed, outlining two stages: (1) the decision to maintain EU market access through internal policy alignment, and (2) alignment with the AI Act outside the EU. The analysis finds strong evidence that OpenAI’s internal alignment was influenced by the EU’s market size and its regulation of inelastic targets, while regulatory capacity provided a limited explanation. For external alignment, moderate support was found for the impact of the AI Act’s stringency and the non-divisibility of its targets. While the Brussels Effect explains much of OpenAI’s alignment with the AI Act, this study acknowledges its limitations and the potential role of alternative explanations. As one of the first studies to empirically assess the Brussels Effect in relation to the AI Act, this research contributes to ongoing debates about the Brussels Effect and the EU’s ability to shape global norms.