The EU AI Act: a medley of product safety and fundamental rights?

Abstract

The European Union (“EU”) draft for an Artificial Intelligence Act (the AI Act) is a legal medley. Under the banner of risk-based regulation, the AI Act combines two repertoires of European Union (EU) law, namely product safety and fundamental rights protection. However, the proposed medley can fail if it does not account for the structural differences between the two legal repertoires. This paper maps how three classical issues of law and technology—the pacing problem, a mismatch between means and ends, and institutional path dependence—manifest themselves in the AI Act. After this diagnosis, it proposes some adjustments to the text and spirit of the AI Act.

Keywords: risk regulation, product safety, fundamental rights, harmonized standards, legitimacy