Why the AI Act Won’t Trigger a Brussels Effect

Abstract

Scholars have widely discussed whether the Artificial Intelligence (AI) Act of EU law will trigger a new Brussels effect, namely, the power that EU law exerts beyond its own boundaries and jurisdiction in such fields of regulation as data protection, environmental law, or antitrust. The paper argues that both exogenous and endogenous reasons suggest that this will not be the case with the AI Act. In addition to competition among legal systems, e.g., U.S. law, several crucial limits in the normative design of the AI Act support the claim. The result will rather be a patchwork effect. On the one hand, some pieces of the EU legislation on bans of technology and high-risk uses of AI can affect other jurisdictions and the private sector; yet, on the other hand, the troubles with the overall architecture of the regulation will make it unexportable with all its dichotomies and corresponding drawbacks.

Keywords: Artificial Intelligence, Brussels Effect, Governance, Legal Regulation, Neutrality Principle, Risk

JEL Classification: K10, K33