Privacy advocates sound warning on Toronto police use of facial recognition technology

Toronto Police say a pilot project using facial recognition technology has been effective in helping investigators solve crimes, but privacy and civil rights advocates are concerned by a lack of clear guidelines for its use.

Facial recognition pilot project was an effective tool for investigators, Toronto police say

In this Oct. 31, 2018 file photo, a man who declined to be identified has his face painted to represent efforts to defeat facial recognition technology. (Elaine Thompson/Associated Press)

For more than a year, Toronto Police have been using facial recognition technology without widespread public knowledge and while it could be an effective crime fighting tool, privacy advocates are concerned the risks to civil liberties outweigh the benefits.

Deputy Chief James Ramer tried to demystify its use at a Toronto Police Services Board meeting Thursday, where police presented their report on the facial recognition pilot project they ran last year.

"I would liken this technology to a virtual mug shot book, rather than binders and binders of photographs that someone would be looking through," said Ramer, adding that the system used during the pilot project ensured witnesses and victims of crime would not be re-traumatized.

"This process is much more compassionate to many of the victims of the various types of assault who previously would have to go through mug shot books and photo lineups."

Toronto police spent $451,718 to purchase the system in March 2018, using funding from a provincial Policing Effectiveness and Modernization Grant.

Deputy Chief James Ramer says the one-year pilot program using facial recognition simply ensured witnesses and victims of crime did not have to be retraumatized. (Martin Trainor/CBC)

The pilot project ran from March 22, 2018 to Dec. 31, 2018 and according to Thursday's report, it was an "immediate success" leading to the identification of previously unknown criminal suspects.

Tell us what you think!

Help shape the future of CBC article pages by taking a quick survey.

PIctures and still images from surveillance video or witnesses were compared against the Toronto Police Service's database of approximately 1.5 million mugshot images.

The system was able to generate potential matches for about 60 per cent of those images.

Police say about 80 per cent of those matches led to the identification of criminal offenders and follow-up investigations led to their arrests and prosecutions, including four homicides, multiple sexual assaults, a large number of armed robberies and numerous shootings and gang related crimes.

Michael Bryant says the Canadian Civil Liberties Association wants an immediate moratorium on the use of facial recognition technology. Studies have shown that such systems have a higher rate of false positives for black faces than white faces. (CBC)

But that doesn't carry much weight with privacy and civil liberties advocates.

"Nobody knew about it. Nobody knew it was being used and how it was being used," said Michael Bryant of the Canadian Civil Liberties Association.

He called for an immediate moratorium on the use of facial recognition technology until standards, checks and balances can be developed and debated by city council.

"Facial recognition technology as it is being used today is carding by algorithm and by a notoriously unreliable algorithm at that," said Bryant.

There have been concerns that such systems have a racial bias as false positives for black faces are higher than for white faces.

Ann Cavoukian, the former three-time Information and Privacy Commissioner of Ontario, says the lack of debate about the rollout of this technology is a big concern.

Ann Cavoukian, former privacy commissioner of Ontario, says the Toronto Police Service use of facial recognition software lacks transparency and could threaten an individual's right to privacy. (Paul Smith/CBC)

"There is no transparency associated with this," Cavoukian said.

" And you can't hold people accountable if you don't know what's going on. So this has been taking place for a year. And while people may not be aware of it, there's a very high false-positive rate for facial recognition".

She says there is concern the technology's use in public spaces presents a threat to individual privacy and could be abused.

But Ramer says Toronto police have no intention of using the system to randomly scan faces in public places.

"The Crown has made it very clear that the use of facial recognition in any other way as other jurisdictions have done in scanning crowds would require judicial authorization," he said.

"We have never signed such authorization because we do not use facial recognition in that way."

Still, Cavoukian is concerned that no guidelines or oversight has been put in place.

"Have independent oversight and you know it can't just be sprung on us a year after the fact."