AI Federalism: Shaping AI Policy within States in Germany

Abstract: Recent AI governance research has focused heavily on the analysis of strategy papers and ethics guidelines for AI published by national governments and international bodies. Meanwhile, subnational institutions have also published documents on Artificial Intelligence, yet these have been largely absent from policy analyses. This is surprising because AI is connected to many policy areas, such as economic or research policy, where the competences are already distributed between the national and subnational level. To better understand the current dynamics of AI governance, it is essential to consider the context of policy making beyond the federal government. Although AI may be considered a new policy field, it is created, contested and ultimately shaped within existing political structures and dynamics. We therefore argue that more attention should be dedicated to subnational efforts to shape AI and present initial findings from our case study of Germany. Analyzing AI as a policy field on different levels of government will contribute to a better understanding of the developments and implementations of AI strategies in different national contexts.