Navigation

© Zeal News Africa

Controversial AI Body Cameras Debut in Canada: Facial Recognition on 'Watch Lists' Sparks Debate

Published 1 hour ago4 minute read
Uche Emeka
Uche Emeka
Controversial AI Body Cameras Debut in Canada: Facial Recognition on 'Watch Lists' Sparks Debate

A contentious pilot project has commenced in Edmonton, Canada, utilizing police body cameras equipped with artificial intelligence-driven facial recognition technology. The system is designed to detect the faces of approximately 7,000 individuals included on a “high risk” watch list, marking a live test of technology that has been widely criticized for its intrusive nature within North American policing. This initiative, led by Axon Enterprise, Inc., a prominent manufacturer of body cameras, has ignited widespread alarm. Notably, Axon had previously acknowledged serious ethical concerns regarding police use of facial recognition technology six years prior, leading them to temporarily abandon its development in 2019.

Critics, including Barry Friedman, former chair of Axon’s AI ethics board and now a law professor at New York University, have voiced profound concerns. Friedman asserts that Axon is proceeding without sufficient public discourse, rigorous testing, or expert scrutiny regarding the significant societal risks and privacy implications. He emphasizes the necessity of a clear demonstration of benefits before deploying technologies with such real costs and risks. In contrast, Axon’s founder and CEO, Rick Smith, characterizes the Edmonton pilot not as a product launch but as “early-stage field research.” This research aims to evaluate the technology’s performance, identify necessary safeguards for responsible use, and gather independent insights from real-world conditions outside the U.S. before potential wider implementation, including within the United States.

The Edmonton Police Service (EPS) acting superintendent, Kurt Martin, states that the pilot's primary objective is to enhance officer safety. This is achieved by enabling body-worn cameras to identify individuals classified with a “flag or caution” for categories such as “violent or assaultive; armed and dangerous; weapons; escape risk; and high-risk offender.” As of December 2, 2024, this watch list encompasses 6,341 individuals, supplemented by a separate list of 724 people with serious criminal warrants. Ann-Li Cooke, Axon’s director of responsible AI, reiterated that the targeting is specific to individuals with serious offenses. Axon, a publicly traded company renowned for the Taser, dominates the U.S. body camera market and is expanding its reach internationally. This pilot could significantly influence policing globally. While competitor Motorola Solutions also possesses the capability to integrate facial recognition into body cameras, it has deliberately refrained from deploying it for proactive identification based on its ethical principles, though not ruling out future use. The government of Alberta had previously mandated body cameras for all provincial police agencies in 2023, framing it as a transparency measure.

The prospect of real-time facial recognition identifying individuals in public spaces has faced considerable backlash. In the U.S., it has been unpopular across the political spectrum, with civil liberties advocates and discussions on racial injustice prompting Axon and other tech giants to halt sales of such software to police. Major concerns stem from studies revealing that the technology often produces biased results concerning race, gender, and age, and performs less accurately on real-time video feeds compared to posed identification photos. Several U.S. states and numerous cities have sought to restrict police use of facial recognition, although the Trump administration has attempted to deter such state-level regulations. Internationally, the European Union has banned real-time public face-scanning police technology, with exceptions for serious crimes. However, the United Kingdom, no longer part of the EU, initiated testing a decade ago and has utilized it for 1,300 arrests in the last two years, contemplating national expansion.

Many operational specifics of Edmonton’s pilot remain undisclosed. Axon utilizes a third-party vendor for its AI facial recognition model but declined to name the provider. The EPS has confirmed the pilot will run until the end of December and only during daylight hours, citing factors like lighting conditions and cold winter temperatures as critical considerations for a successful proof of concept. Approximately 50 officers involved in the pilot will not receive real-time match notifications; outputs will be analyzed post-shift at the station. In the future, the technology could potentially alert officers to dangerous individuals during investigations or responses to calls, rather than during routine patrols. Officers can switch cameras to a higher-resolution active recording mode when responding to calls. The office of Alberta’s information and privacy commissioner is currently reviewing a privacy impact assessment submitted by the Edmonton police, a prerequisite for projects handling “high sensitivity” personal data.

University of Alberta criminology professor Temitope Oriola views Edmonton as a

Recommended Articles

Loading...

You may also like...