A new report has revealed that the Toronto police were using Clearview AI’s software to advance their investigations, even when their superiors did not know that they had access to the technology. The report comes courtesy of CBC News, which obtained an internal document detailing the department’s use of Clearview AI through an access to information request.
The Toronto police had previously acknowledged that officers had used Clearview AI on a trial basis between October 2019 and February 2020. Mark Saunders, the police chief at the time, would eventually order the department to stop using the technology on February 5, shortly after the January New York Times report that first brought Clearview’s activities to light.
The CBC report details the scope of the Toronto police’s facial recognition program. Clearview was ultimately used by 115 individual officers, who collectively uploaded more than 2,800 phots (and conducted more than 2,800 facial recognition searches) with the Clearview database. Searches were conducted over the course of 84 active investigations, 25 of which were advanced as a direct result of the use of facial recognition. In some cases, the police used the technology to identify a suspect. In others, it was used to identify a victim or a witness.
The problem, according to privacy watchdogs, is that those officers never had formal permission to perform any of those searches. Clearview’s technology first cropped up on the Toronto police’s radar when a detective saw a demo at a victim identification conference in the Netherlands in October of 2019. Clearview gave Toronto officers access to a free trial shortly thereafter (the company made free trials readily available to police as part of a broader marketing strategy). It then spread rapidly within the department, with officers hosting an internal Clearview showcase for around 100 investigators in December of that year.
Unfortunately, none of the officers using the technology seem to have stopped to consider whether or not it was legal before using it, or before sharing it with their peers. Officers did not bother to consult with Crown attorneys until January, after the New York Times report called its legality into question. Saunders himself does not seem to have been aware that his own department was using Clearview at that time, since investigators only discussed the technology with their own internal supervisors, and with Clearview itself.
The Toronto police could end up paying for that lack of foresight in court. The department is currently handling two cases that were put together with evidence from Clearview. However, four Canadian privacy commissioners (including the federal privacy commissioner) have ruled that Clearview’s tech violates Canadian privacy law, which means that those cases could be thrown out due to police violations of people’s civil liberties.
Clearview abandoned the Canadian market in the summer of 2020, when the commissioners opened their investigations. The Toronto police stressed that they do not have any plans to partner with Clearview in the future, and indicated that the technology would be deemed an extreme compliance risk under its new vetting system for new technologies.
Source: CBC News
January 7, 2022 – by Eric Weiss