To comb through the thousands of hours of body camera footage accumulated by police departments, some agencies are turning to artificial intelligence for help — but the task is not proving to be a simple one.
In recent years, body-worn cameras have been widely adopted by police departments across the country in an effort to ensure transparency and accountability within law enforcement and to restore public trust. However, a recent investigation by ProPublica revealed that the impact of body-worn cameras has been fairly limited. According to the report, the sheer volume of footage generated by these body cameras has posed a significant challenge for police departments.
Axon, the nation’s largest provider of police cameras and cloud storage for video footage, reported a database size exceeding 100 petabytes, equivalent to storing more than 5,000 years of high-definition video. But despite this vast repository of data, due to resource limitations a significant portion remains unexamined, merely stored away and rarely reviewed.
Jonathan Wender, head of Polis Solutions, a company offering analytic tools powered by artificial intelligence to address this data problem, shared his thoughts on the issue with ProPublica.
“In any community, body-worn camera footage is the largest source of data on police–community interactions. Almost nothing is done with it,” he said.
In response to this challenge, some police departments are turning to new technologies to analyze the accumulated footage for insights into officer behavior and patterns. One such agency is the Paterson Police Department in New Jersey, which has adopted analytic tools from Truleo, a Chicago-based software company.
Chief Isa Abbassi, tasked with reforming the department, highlighted the importance of technology in rebuilding public trust.
“Changes in Paterson are led through the use of technology,” Abbassi stated in a press conference.
The Ann Arbor Police Department in Michigan also partnered with the AI company.
Truleo’s software, which costs approximately $50,000 annually, enables supervisors to flag specific behaviors exhibited by officers during interactions captured on body cam footage.
“There are certain officers who don’t introduce themselves, they interrupt people, and they don’t give explanations. That officer’s headed down the wrong path,” Anthony Tassone, Truleo’s co-founder, said.
Other companies such as Polis Solutions, as well as academic institutions like the University of Southern California and Washington State University, are developing AI-powered tools to analyze body camera footage and identify trends in officer behavior.
Despite these technological advancements, concerns remain regarding the transparency and accountability of policing practices.
Christopher J. Schneider, a professor at Brandon University, was skeptical about the effectiveness of AI tools in addressing systemic issues within law enforcement. He also raised concerns about the confidentiality surrounding the findings of these analyses.
“Even if police departments buy the software and find problematic officers or patterns of behavior, those findings might be kept from the public just as many internal investigations are,” Schneider said.
Moreover, the lack of disclosure surrounding the use of AI tools in policing raises questions about accountability. For instance, some police departments, including Philadelphia’s, have policies that restrict disciplining officers based on spot-check reviews of video footage, a move criticized by civilian oversight bodies like the Police Advisory Committee.
