The system—which cost $451,718 plus annual maintenance and support fees—takes images of potential suspects found on public or private cameras and compares them to an internal database of around 1.5 million mugshots, according to a report submitted in advance of a Thursday board meeting. Between March and December of 2018, police ran 1,516 searches using some 5,000 still and video images; the system generated potential mugshot matches for about 60 per cent of them. Around 80 per cent of the matches led to criminal offenders being identified. (Toronto Star)

Talking point: Police said they purchased the system to identify suspects more efficiently using fewer resources. The agency said it currently only plans to use the tech to compare images with pre-existing mugshots, and that it doesn’t perform real-time facial recognition. The technology has been criticized for disproportionately misidentifying women and people of colour. Earlier in May, San Francisco banned police and other agencies from using it, saying it could potentially be misused. Though Toronto city officials haven’t officially explored banning the tool, federal members of Parliament are looking into the San Francisco ban to see if something similar should be passed in Canada. Charlie Angus, NDP ethics critic, called for the use of facial recognition to be paused, saying, “This technology is being put in place without any legislative oversight.”