AI, Policy, and Power

• In Event: AI, Governance, and Power

Abstract

This paper examines AI governance using a lens of power. Based on the management and design ethics literature, and the policy literature, respectively, we articulate two typologies of AI actors – those involved in the AI design process and in the AI policy process. Drawing from prominent theoretical frameworks of power, we discuss how direct, indirect, and symbolic dimensions of power are manifesting in these two networks of actors. By synthesizing what scholarship has suggested regarding the hypothetical channels of power between actors, and whether and how that power is exercised in practice, we identify key learnings and interrogate the gaps and neglected areas. Based on these insights, we discuss how key AI ethical goals are being translated into action in the private and public sectors, and in the process being compromised. This review helps map the power balance (or imbalance) between actors in AI governance, and suggests remedies for course correction.

Authors

• Indira Patil, Purdue University

• Daniel S. Schiff, Purdue University