Citizen Lab認証済みアカウント

Learning from these lessons, Canada has the opportunity to adopt human rights-focused criminal justice reform. Today, the Citizen Lab and @IHRP_UofT published a first-of-its-kind report that focuses on the human rights impacts of algorithmic policing technology in Canada.

In 2020, the Chicago Police Department decommissioned a 10-year experimental predictive policing program. The reasons for cancelling Chicago’s program included the “unreliability” of the technology & “negative consequences” related to arrests that never resulted in convictions.

In 2020, the Chief of Police in Detroit reported that facial recognition technology fails to correctly identify people “96% of the time.”

But there are more important questions than cost. The human rights dangers surrounding the technology are glaring given their susceptibility to bias and inaccuracy. And lessons have been learned in the United States.

Creating and maintaining police algorithms is resource intensive, requiring publicly-funded human capital, tech support, and/or commercial licensing fees.

These are only examples among many of the algorithmic policing technologies under development or in use in Canada. But the true cost of experimentation with algorithmic policing is probably much higher.

And the York Regional Police Service budgeted $1.68 million in 2019 to pay for a facial recognition system.

The City of Calgary reportedly paid $1.4 million to Palantir Technologies Inc. to give the Calgary Police service over three years—a figure which doesn’t include amounts paid to other technology vendors that the police service is also known to be using.

The Saskatchewan Police Predictive Analytics Lab was reportedly expected to receive nearly $1 million dollars over two years for their predictive policing program in Saskatoon.

Accurate figures about how much money Canadian police services have poured into experimentation with algorithmic technology are not fully known. But some details are available:

In Los Angeles, one of the first cities to become a testing ground for predictive policing technology, the city reportedly spent over $17 million USD between 2010-2015 on the program.

They're not designed to redress the root causes of crime. This alone is reason to abandon public investments in such technology. But at a time when police budgets in Canada are being closely re-examined, understanding the financial costs of this tech takes on new importance.

Researchers are sounding alarm bells about its human rights dangers. It has already led to wrongful arrests of racialized individuals in known cases. Algorithms tend to replicate existing patterns of policing, including racial disparities in policing.

Police services around the world—and now in Canada—are starting to use algorithmic technology in ways that the public has not seen before. Algorithmic policing technology is controversial. A thread on the true cost of this tech

As a result, these communities are the most likely to be negatively impacted by the use of algorithmic policing technologies, and are at the greatest risk of having their constitutional and human rights violated.

These technologies’ algorithms are generally trained on biased historical police data that include anti-Black racism, anti-Indigenous racism, and discrimination against racialized individuals and communities more broadly.

In collaboration with @IHRP_UofT, this report examines predictive policing and algorithmic surveillance technologies used by Canadian police services, concluding that they present several human rights and constitutional law issues.