European standardization in flux: Navigating delegation and control in AI governance

The EU confronts a dilemma in AI governance: on the one hand, it faces strong incentives to delegate detailed, technical rulemaking to private authorities. On the other hand, the political charge of these detailed rules makes it imperative to retain control over them. We investigate how the EU navigates this dilemma, and which tactics are available to it to control delegated governance in the case of AI standardization. Building on theories of global governance (Bernstein, 2010; Ruggie, 2014), we reveal the hard and soft mechanisms the European Commission can use to influence both the operation and the output of standards development organizations (SDOs). We argue that the EU’s control tactics depend on (1) its relationship with the SDO in question and (2) whether standard setters serve through delegation with granted authority or orchestration with enlisted authority. The EU is often portrayed as lagging rival superpowers, not only in AI development but also in ICT standardization. As our analysis demonstrates, however, the EU has a range of robust tactics to interfere in technical standard-setting, but in opting to deploy these tactics, might risk the overall integrity and legitimacy of European standardization more broadly.