Montreal Police AI Cameras Spark Privacy Backlash

Montreal police are expanding their use of AI-powered surveillance tools to track vehicles and people across the city, and the deployment came to light only after a reporter happened to spot one of the cameras mounted on a lamp post, raising real questions about transparency around how the technology is actually being rolled out, according to CBC News's reporting on the discovery.

A Radio-Canada reporter spotted a high-powered camera capable of scanning and identifying vehicles on Henri-Bourassa Boulevard in Montreal's Rivière-des-Prairies neighbourhood earlier this month. The device, later confirmed by the Service de police de la Ville de Montréal, was a Cloudrunner CR-H2, manufactured by Montreal-based company Genetec. Within days of Radio-Canada publishing its story about the camera, it was taken down, according to CTV News's coverage of the removal.

The Cloudrunner system does considerably more than read a license plate number. Similar to the controversial Flock Safety cameras widely deployed across the United States, the system also captures and stores specific vehicle characteristics, including colour, make, and model, according to the manufacturer's own website, cited in CBC News's reporting. That expanded data collection is exactly what's fueled controversy around comparable systems south of the border, where Flock cameras have become a genuine flashpoint in cities weighing crime-solving benefits against surveillance concerns.

Genetec vice-president Andrew Elvish defended the technology's design while acknowledging the genuine tradeoff involved. "You should have an expectation of reasonable privacy," he said, before adding, "it's a trade off. If the threat of something bad happening is high, as a community, what are we comfortable with giving up to ensure that the community is safe?"

This single camera appears to be just one piece of a wider pattern. The SPVM separately confirmed it is using iMotion ROC AI software, developed by Rank One Computing, to help investigators sift through large volumes of video footage. SPVM spokesperson Anik de Repentigny was specific about the software's stated limits: "This software will not be used to verify the identity of individuals in video images, but rather to identify the video passages where the targeted known element appears," she said in a statement to CBC News, explicitly denying the force is using the software's facial recognition or biometric capabilities.

Notably, the City of Montreal itself distanced from the deployment when asked. A city spokesperson said it had not installed the Cloudrunner camera, referring all questions about its deployment directly to the SPVM, according to CBC News's reporting, a response that highlights genuine ambiguity about which body actually holds accountability for this kind of technology rollout.

This tension isn't unique to Montreal. In Massachusetts, where Flock Safety technology now operates in more than 80 cities and towns, Quincy Police Chief Mark Kennedy has credited automatic license plate readers with tangible investigative results, including clearing more than 40% of motor vehicle hit-and-runs compared to a national average of roughly 8% to 10%, according to Yahoo News's reporting on the technology's use there. Kennedy directly credited the cameras with helping solve shootings, stabbings, and sexual assaults.

But the same underlying capability has triggered pushback elsewhere. Denver's city council rejected a $666,000 extension to its license plate reader contract with Flock Group, with council members specifically citing concerns that gathered data could be shared with other agencies, including those working closely with federal immigration authorities, according to Axios's reporting on that decision. That specific concern, data sharing beyond its original stated purpose, is exactly the kind of transparency question Montreal's own unexplained deployment has now raised locally.

This case is worth understanding for any business operating in public safety technology, municipal contracting, or data governance more broadly. It illustrates a genuine reputational risk pattern that keeps repeating across AI surveillance deployments: technology installed quietly, without proactive public disclosure, tends to generate significantly more backlash once discovered than if the same rollout had been announced transparently from the start.

For companies selling AI surveillance or monitoring technology to public sector clients, Montreal's rapid camera removal following media scrutiny is a clear signal that public agencies remain highly sensitive to transparency criticism, and vendors should factor proactive disclosure planning directly into deployment strategy rather than treating it as the client's problem alone.

What is the Cloudrunner camera Montreal police used?
The Cloudrunner CR-H2 is an AI-powered license plate recognition camera made by Montreal-based company Genetec, capable of scanning license plates and capturing additional vehicle details including colour, make, and model.

Why was the Montreal camera taken down?
The camera was removed within days after Radio-Canada published a story revealing its presence on a lamp post in the Rivière-des-Prairies neighbourhood, following public attention to its previously undisclosed deployment.

Is Montreal police using facial recognition technology?
The SPVM has said its iMotion ROC AI software, used to help investigators search through video footage, does not use facial recognition or biometric identification capabilities, and is instead used to locate video segments containing a specific known element.

Montreal police quietly deployed an AI-powered license plate camera that was discovered by a reporter and removed within days of media coverage, raising questions about transparency in the force's broader AI surveillance rollout. The technology, similar to controversial Flock Safety cameras used in the US, captures vehicle characteristics beyond just license plate numbers. Similar AI license plate reading technology has produced measurable crime-solving results in some U.S. cities while triggering rejected contracts and privacy pushback in others, reflecting an unresolved tension between investigative value and surveillance concerns.