Politics and policy of Artificial Intelligence

This paper aims to examine state of the art and to set out a research agenda to examine politics and policy of Artificial Intelligence (AI). Social studies of AI so far have mostly focused on ethical, philosophical, legal and economic issues, with politics and policy of AI having received relatively less attention. To address this gap, this paper aims to integrate insights and concepts from policy studies, Political Science, and Science and Technology Studies to address political, policy and power aspects of AI. In particular, it aims to explore five topics. First, how AI and the concept of technology more broadly is understood and defined in politics and policy? Do instrumental understanding of technology as a tool to pursue economic and social aims dominate in AI policy and politics? Are political, social and cultural aspects of AI and technology recognized as well? Who defines AI and technology and what are broader implications of specific definitions and understandings? Second, what is the role of ideational and cognitive dimension in AI policy? What do ideational approaches such as socio-technical imaginaries, narratives, discourses, frames and policy scripts reveal about AI politics? Third, what manifestations of power do we encounter in AI politics and policies? How is power distributed among different actors such as government, companies and the public or employers and employees? How does power relate to intersectional characteristics such as race, gender or socio-economic status? Fourth, in a fast-developing area of AI policy and politics, what is relationship between change and continuity? Do novel ideas emerge or do we see strong path-dependences of established discourses? Fifth, how do intertwined local, national, regional and global processes play out in the field of AI politics and policy? Do we see policy learning and convergence or do diversity and divergence across national and regional policies dominate? And how do dynamics of global cooperation and competition play out in AI politics and policies?