Edmonton Police Pilot AI Facial Recognition Body Cams, Testing 7,000-Person Watchlist
AI Body Cameras Tested in Edmonton for Facial Recognition

In a landmark move that could set a precedent for law enforcement in North America, the Edmonton Police Service has initiated a live pilot program deploying body cameras integrated with facial recognition artificial intelligence. This significant test involves scanning for individuals on a curated watchlist, marking a major strategic shift in the use of surveillance technology on the streets.

The Edmonton Pilot: Details and Scope

The pilot program, which commenced last week, involves approximately 50 officers in the city of Edmonton, home to over a million residents. The core function of the technology is to scan and identify faces from a predefined list of about 7,000 individuals deemed "high risk." This list, as detailed by acting superintendent Kurt Martin, comprises 6,341 high-risk offenders and an additional 724 people with serious criminal warrants outstanding.

These individuals are categorized for reasons including violent offenses, weapons possession, being an escape risk, or having serious warrants. However, during this initial "proof of concept" phase running until December, officers will not receive immediate, real-time alerts on their devices. Instead, all data captured by the cameras will be analyzed retrospectively back at the station.

Rick Smith, the CEO of Axon Enterprise, Inc.—the company providing the technology—clarified that this is not a commercial product launch. He described it as "early-stage field research" aimed at testing the system's performance in genuine operational environments, including the challenges posed by Edmonton's limited winter daylight and cold temperatures.

Ethical Firestorm and Technological Caution

The deployment has ignited immediate criticism from privacy advocates and former advisors to Axon itself. Barry Friedman, a New York University law professor and former chair of Axon's own AI ethics board, expressed strong concerns to The Associated Press. He argued the company is moving forward without adequate public discourse or expert review, stating it's crucial not to deploy such high-risk technologies without clear evidence of their benefits. Friedman had resigned from the board in 2022 following a prior disagreement over Taser-equipped drones.

Axon itself has acknowledged the existing limitations of facial recognition AI. The company admitted to the AP that factors like distance, lighting, and camera angle can still hinder accuracy and that these issues may "disproportionately impact accuracy for darker-skinned individuals." Emphasizing a human-in-the-loop approach, Axon stated that every potential match flagged by the AI must undergo verification by a human officer.

Global Context and Community Impact

This pilot emerges amid a fragmented global landscape for facial recognition. While the European Union has enacted broad bans on real-time public face scanning, countries like the United Kingdom have used similar technology to make arrests. In the United States, Axon's main competitors, such as Motorola Solutions, have refrained from real-time identification on ethical grounds.

For Axon, the dominant U.S. supplier of police body cameras, the Edmonton pilot serves as a critical data-gathering exercise. The insights gained are intended to help build stronger oversight frameworks for potential future deployment in American jurisdictions.

Closer to home, Temitope Oriola, a criminology professor at the University of Alberta, called Edmonton a "laboratory" for this tool. He pointed to the police service's historically complex relationship with Indigenous and Black communities, noting it is yet to be determined whether this technology will genuinely enhance public safety or further exacerbate existing tensions and mistrust.