Between Fundamental Rights and Existential Risks: Civil Society and the EU AI Act

This article examines how Civil Society Organisations (CSOs) articulated their positions in the policy-making process for the EU AI Act. Rather than treating “civil society” as a unified bloc, this article shows how civil society actors themselves compete to define what counts as risks, harms and appropriate regulation in EU AI governance. Utilising a discursive framing approach, the article reconstructs the arguments of two highly active organisations that exemplify a fundamental cleavage within civil society: while traditional digital civil rights groups highlight fundamental rights as the most pressing issue to be dealt with in the AI Act, more recently founded, US-based groups with a specific AI-focus primarily advocate for the mitigation of existential and systemic risks posed by AI systems. As space for civil society influence in EU governance is sparse, which makes it likely that one of these risk conceptions is going to win out over the other, mapping these competing framings is essential to understanding the structure and contested nature of the policy space around the making and implementation of the AI Act.