Search
Program Calendar
Browse By Day
Browse By Room
Search Tips
Virtual Exhibit Hall
Personal Schedule
Sign In
As AI technologies rapidly advance, especially in public sectors like policing, it is essential to reconsider how novel technologies reshape public organizations, and importantly the practices’ of actors involved in policy implementation through empirical research. This paper challenges traditional views on bureaucracy and builds on the concept of algocracy by introducing AI-level bureaucracy and a new organizational actor involved in the making of AI-level bureaucracy: dual experts. I specifically conceptualize AI-level bureaucrats as public agents – both human and AI – who exercise discretion at the "street level", in direct contact with the public, in implementing public policy. Additionally, I illustrate how this form of bureaucracy occurs in practice based on extensive field research in six police organizations with insights from over 115 law enforcement actors through three empirical examples: hotspot community policing, gender/domestic violence risk assessment tools, and AI avatars.
In this paper, I argue that discretion has not disappeared in the AI age, it has shifted and requires a shift in our understanding. Public policy implementation now involves various actors (public and private), and new organizational roles, thus creating new forms of power, with more intricate and sometimes conflicting bureaucratic dynamics. To understand this shift in practice, it is crucial to consider these dynamics as an assemblage, where diverse agents, factors, and forces, that can range from politics, social dynamics, or individual characteristics, contribute to the making of AI-level bureaucracy. For instance, dual experts play an important role in making AI-level bureaucracy, considering they hold both epistemic authority of expertise in tech-related fields (e.g., AI, computer/data science) and experience as public actors (e.g., police office).