Toward an ‘Ever Closer Union’

The Making of AI-Ethics in the EU

Abstract

More than any other region attempting to get ahead in the global AI race, the EU has emphasized a commitment to ‘AI-ethics’ and invested significant work in the development of principles and rules for the ethical governance of AI. This paper examines the production, performance, and politics of AI-ethics in the EU through a close, co-productionist reading of how the “European Group on Ethics in Science and New Technologies” has framed AI and its desirable relationship to humans. Our analysis shows that the making of AI-ethics in this context extends far beyond the settlement of principles and norms for AI. Instead, we argue that AI-ethics is, at the same time, performing authoritative acts of ontological classification that cut the human clean from AI to render it controllable and manageable. These ontological politics, we show, serve to embed AI within long-held imaginaries of European unification beyond market harmonization and re-configure how the EU imagines to achieve an ‘ever closer union’ among its members in the innovation era. Different to attempts at deeper integration through the mobilization of science, the turn to AI-ethics presents a novel rationale through which the EU legitimizes its authority to govern, suggesting a constitutive role for ethics in the EU’s contemporary integration efforts.