Citizen Perceptions of Artificial Intelligence Use in Policing

• In Event: Poster Session: Urban Politics and Policy & Public Policy
  In Poster Session: Poster Session: Frontiers of Public Policy

Abstract

Policing agencies are increasingly adopting advanced technological tools powered by artificial intelligence (AI) to resolve crimes and improve internal operations. Yet growing public and regulatory scrutiny of these tools—and of law enforcement agencies generally—suggests that citizen support and trust will be critical for successful implementation. We administer a survey experiment to assess public attitudes toward AI in policing. First, we consider whether public support varies when AI is used for external aims, such as predictive policing, versus internally-oriented purposes, like automated review of case files to detect misconduct. Next, in line with prior work on preferences for local governance, we examine whether citizens are more supportive of AI tools when the implementing agency is a local sheriff’s office versus a federal agency like the FBI. Finally, given that policing agencies differ in their capacity to adopt advanced technologies, we evaluate whether information about agency capacity influences public perceptions.

We find that conservatives are significantly more likely to support external versus internal applications of AI in policing, although average Democrat support is higher than Republican support for both external and internal uses. Moreover, citizens generally prefer sheriff use over FBI use of AI—they are more willing to trust the implementing agency and are even more willing to pay higher taxes and contribute their own personal data to improve the AI tool. Surprisingly, we find no impact of capacity concerns on public attitudes toward AI in policing. The findings have important implications for public support for police reform.

Authors

• Kaylyn Jackson Schiff, Yale University

• Ian T. Adams, University of utah

• Josh McCrain, University of Utah

• Daniel S. Schiff, Purdue University