Officer Acceptance and Use of AI-Driven Body-Worn Camera Footage Review

Abstract

Police departments review only a fraction of the enormous amount of body-worn camera (BWC) footage recorded daily by their officers. The inability to review BWC footage may undermine many of the perceived benefits of BWCs, most notably increased accountability and civility during officer-community member interactions. Artificial Intelligence (AI) has the potential to overcome this problem through analysis of vast amounts of BWC footage, but AI’s utility hinges on officer acceptance and use of the technology. In the current study, the authors measure attitudes and use of Truleo, an AI-based program that uses natural language processing to deconstruct, analyze, and categorize BWC footage audio. We conducted surveys of officers and supervisors in two medium-sized Arizona police departments both two months before and five months after the departments deployed Truleo via randomized controlled trials. Attitudes about Truleo pre-deployment were mostly neutral, and they remained so after deployment. Post-deployment, officers did report increased understanding of Truleo, and they became less concerned about how the AI platform was being used. In terms of use, officers accessed Truleo 1.6-1.8 times each week, and supervisors averaged 3.6-4.6 times per month. Results suggest AI-driven BWC review can be successfully integrated into a police department’s day-to-day operations.

version: pre-print