Antitrust in Artificial Intelligence Infrastructure: Between Regulation and Innovation in the EU, the US, and China

The enormous amount of data and the substantial computational resources are crucial inputs of artificial intelligence (AI) infrastructure, enabling the development and training AI models. Incumbent firms in adjacent technology markets hold significant advantages in AI development, due to their established large user bases and substantial financial resources. These advantages facilitate the accumulation of enormous amounts of data, and the establishment of computational infrastructure necessary for sufficient data processing and high-performance computing. By controlling data and computational resources, incumbents raise entry barriers, leverage advantages to favour their own AI services, and drive significant vertical integration across the AI supply chain, thereby entrenching their market dominance and shielding themselves from competition. This article examines regulatory responses to these antitrust risks in the European Union (EU), the United States (US), and China, given their leadership in digital regulation and AI development. It demonstrates that the EU’s Digital Markets Act, and China’s Interim Measures for the Management of Generative Artificial Intelligence Services introduce broadly framed yet applicable rules to address challenges related to data and computational resources in AI markets. Conversely, the US lacks both AI regulations and digital-specific competition laws, instead adopting innovation-centric policies aimed at ensuring its AI dominance globally. Given the strategic importance of AI development, all three jurisdictions have adopted a cautious approach in investigating potential abusive practices.