USC research using artificial intelligence to analyze policy officer body cam video may improve officer training. Composite: Letty Avila. Image Source: iStock.)

AI meets policing: USC leads LAPD body cam study

New research using artificial intelligence aims to improve officer interactions with the public.
ByUSC Dornsife News Staff

Why it matters: The Los Angeles Police Department (LAPD) is stepping into the future, leveraging artificial intelligence (AI) to analyze the communication between officers and drivers during traffic stops.

  • The research is led by a multidisciplinary team based at USC, including scholars from psychology, sociology, engineering, computer science, computer science, public health, and public policy.

The big picture: The Los Angeles Times reports that the LAPD’s move to integrate AI into its operations is part of a broader initiative to improve communication between its officers and the public. The study aims to provide clarity and actionable insights.

What’s happening: Over three years, the Everyday Respect research team will:

  • Review body camera footage from roughly 1,000 traffic stops.
  • Develop measurements of interactions between officers and the public based on department policy as well as feedback from community members and police officers.
  • Feed these measures into a machine-learning tool that will autonomously analyze thousands of additional videos.

Between the lines: Findings from the study will be used to help train officers on how best to navigate encounters with the public and “to promote accountability,” according to Cmdr. Marla Ciuffetelli of the LAPD’s Office of Constitutional Policing and Policy.

  • What she said: Machine learning “is in its infancy, but will undoubtedly become a profound element in officer training in the future.”
  • All data regarding individuals featured in the videos will be anonymized in the final report.

The subjective nature of certain standards means there’s room for debate on what the AI deems “appropriate.” The researchers acknowledge these challenges, emphasizing the importance of diverse feedback and rigorous criteria development.

  • Factors such as the location of the stop and the driver’s race, as well as the officer’s rank, age, and experience, will be part of the analysis. Data regarding all individuals will be kept anonymous.

What to watch: The study’s preliminary results are expected a year into the research.

The bottom line: While acknowledging that it’s “hard to speculate what’s going to happen with the research until it’s done,” Ciuffetelli said the findings will eventually be incorporated into the department’s training models.

Key funders: The research is supported by the National Science Foundation CIVIC Program, Microsoft Justice Reform Initiative, Google Award for Inclusion Research and USC Zumberge Interdisciplinary Award.

This story was first reported by Los Angeles Times reporter Libor Jany.