Vulnerability in the EU AI Act: building an interpretation

Abstract

Vulnerability, a key concept in EU law, remains ambiguous in the Artificial Intelligence Act (AIA), complicating its legal application. This paper offers a comprehensive interpretation of vulnerability, exploring its historical context in EU law and analysing its references in the AIA. It proposes a tripartite definition encompassing identity, context, and power imbalances, arguing that vulnerability is a fluid condition shaped by AI deployment and fundamental rights. Aligning with Article 7(2) of the AIA, this interpretation advocates for a broader, adaptive understanding of vulnerability across EU digital regulations, enhancing legal protections for vulnerable individuals in the digital age.