Clearview AI engaged in illegal mass surveillance: privacy commissioners

The facial recognition technology company used by dozens of police departments across the country engaged in illegal mass surveillance before it withdrew itself from the Canadian market over the summer, the federal privacy watchdog declared in a report released Wednesday. The U.S.-based company, Clearview AI, scraped some three billion images of faces and corresponding biometric identifiers off of publicly accessible information on the internet such as social media. Users of the technology -- in Canada mostly police -- could then upload an image to be matched against the database, and then directed toward the original source of images, according to a joint report from the federal Office of the Privacy Commissioner and its provincial counterparts in British Columbia, Alberta, and Quebec. "What Clearview does is mass surveillance and it is illegal," federal privacy commissioner Daniel Therrien said in a Wednesday statement. "It is completely unacceptable for millions of people who will never be implicated in any crime to find themselves continually in a police lineup." The OPC launched a pair of investigations into the technology and the RCMP back in February after media reports of the technology's use by police departments. Wednesday's report only dealt with the private company, and not the RCMP's role. In a statement emailed to The Wire Report, the RCMP said it would "continue to work closely with the Office of the Privacy Commissioner on its investigation into the RCMP’s use of Clearview AI." "With the nature of crime constantly evolving, law enforcement agencies, including the RCMP, must continuously monitor emerging and evolving technologies. We also recognize that privacy is paramount and a reasonable expectation of Canadians, which must be balanced with the ability of law enforcement to conduct investigations and protect the safety and security of Canadians," an RCMP spokesperson wrote. In its correspondence with the four privacy commissioners over the course of the investigation, Clearview AI disagreed that the commissioners had any jurisdiction over the company, arguing that “none of Clearview’s activities take place in Canada” and that the company “is of the view in the circumstances that none of the statutes invoked apply and that no connecting factors create a real and substantial link to Canada,” according to quotes in the OPC report. In a letter of intention sent to the company, the OPC notified Clearview that it could order or recommend that the company stop offering the facial recognition service; cease the collection, use, and disclosure of the data; and delete the images and biometric facial arrays it had collected from Canadians. The company stopped offering its facial recognition technology in July, but was hesitant to commit to deleting any of the data it collected, writing that it "would be willing to take steps, on a best efforts and without prejudice basis, to try to limit the collection and distribution of the images that it is able to identify as Canadian.” The OPC also left the door open to court action, writing in the report that should Clearview AI refuse to implement any of its recommendations or those of its provincial counterparts, "we will pursue other actions available to us under our respective Acts to bring Clearview into compliance with federal and provincial privacy laws applicable to the private sector." Clearview AI's rejection of the OPC's jurisdiction marks the third time in recent years that high-profile tech companies have rejected the privacy watchdog's recommendations. In October, the OPC found that the use of facial recognition in shopping malls owned by Cadillac Fairview violated Canadian privacy law, but the company refused the watchdog's request for compliance. In 2019, Facebook Inc. also rejected requests from the OPC to comply with federal privacy law in the wake of the Cambridge Analytica scandal, which saw the OPC square off against the social media giant in federal court this month. Information and privacy commissioner for British Columbia Michael McEvoy said at a press conference that Clearview AI's "blatant disregard for Canadian privacy law" was indicative of the need for stronger privacy legislation. But Canada research chair in information law and policy at the University of Ottawa Teresa Scassa said in a phone interview with The Wire Report that it's possible that the kind of activities Clearview AI was engaged in would not be subject to fines under Bill C-11, the federal government's planned privacy law update. The bill calls for the creation of a pseudo-judicial body Personal Information and Data Protection Tribunal, which would ultimately decide on a monetary penalty and to which the OPC could make recommendations. However, Scassa said, "under C-11, there is the power to issue these very substantial monetary penalties, but they are limited to only certain breaches of the legislation." Cleaview AI's "collection and use of personal information without consent: that's not one of them. It's not in the list" of specific breaches enumerated in the bill, she said. Scassa also pointed to an argument made by Clearview AI in its correspondence with the OPC, that the data the company scraped from social media to create its database is exempt from the consent obligation under federal law. "With respect to the consent obligation under federal and provincial legislation … Clearview submits that the exception for publications which are publicly available applies. Information collected by Clearview is nothing more than information available to the public,” Clearview wrote to the OPC, arguing that social media websites are effectively "publications." Scassa said that under C-11, the same exemption exists. "There will need to be new regulations, and you can be sure that industry is making its case to government social media sites have to be included as publications for publicly accessible information," Scassa said. "This is on the table for those regulations. It's a huge issue," she said. "This case shows how much of an issue it is, because this is a company that has built its entire business model on scraping data from social media sites, and then building a facial recognition database based on it, and then saying 'but this is publicly available information.'" "That's going to be a really important conversation, and it’s going to have a huge impact on privacy rights of Canadians." — Reporting by Michael Lee-Murphy at mleemurphy@thewirereport.ca and editing by Anja Karadeglija at akarad@thewirereport.ca

Clearview AI engaged in illegal mass surveillance: privacy commissioners

News | February 3, 2021

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us

Clearview AI engaged in illegal mass surveillance: privacy commissioners

News | February 3, 2021

Unlock all the Canadian telecom, broadcasting and digital media news you need.

FREE TRIAL

SUBSCRIBE

Related Stories

ISED launches consultation on AI computing power

Doing good means doing well financially, Entwistle tells Canadian Club

International cooperation needed to tackle uncertain future of AI, says expert report

Irekvist promotes Ericsson’s efforts to reach net-zero emissions

Reuse & Permissions

Inside The Wire Report

Information

Hill Times Publishing

Get in Touch With Us