Regulating AI Through Standards and the Risk of Preempting Liability: A Transatlantic Comparison

Abstract This article discusses the consequences of the Artificial Intelligence Act adopted by the European Union in 2024 (AIA) on the chance of affirming ex-post liability for damages. As the article concludes, before an ex-ante regulation, such as the AIA, and in the absence of a harmonized liability framework for AI systems, there is a serious possibility of making ex-post liability for damages deriving from AI systems far more problematic to affirm, with the effect, in many cases, of sheltering defendants from the consequences of their acts. In contrast with the often-repeated absence of an exemption from the law in the book in the field of AI, compliance with existing regulatory regimes may preempt tort law, and the unintended consequence of the AIA would be creating a safe harbour for AI.