Labour Rights and the EU Artificial Intelligence Act: How to Get Away with High-Risk AI

Abstract

The European Union’s Artificial Intelligence Act (AI Act) establishes a regulatory framework that primarily approaches worker protection through the lens of employer obligations in the AI supply chain. This chapter critically examines this approach by analysing employers’ legal responsibilities when deploying high-risk AI systems for workforce management, particularly their role as providers and deployers of these systems.

A central focus is the examination of article 6(3) derogations and their implication for classifying high-risk AI systems in workplace contexts. Drawing on interdisciplinary research on workplaces utilising AI automation tools, the chapter identifies potential regulatory gaps in the AI Act’s effectiveness. An examination of article 6’s legislative history reveals that the final text of article 6(3) includes safeguards that mitigate the large-scale classification of high-risk AI, though several challenges remain.

The analysis shows the fundamental limitations in the AI Act’s approach to workplace AI regulation. Although the AI Act establishes employer obligations, its heavy reliance on preconditions and product safety-oriented framework fails to adequately address the power asymmetries inherent in employment relationships. The chapter concludes by advocating for supplementary sector-specific policies to better regulate AI-driven workplace management and protect workers’ interests.

Keywords: EU AI Act, high-risk AI systems, labour rights, legal derogation

JEL Classification: K31