National security agencies' relationship with racialized communities marred by a 'trust gap:' report
External body says AI and the rise of data intelligence could pose a threat to 'vulnerable' people
The relationship between "racialized" groups and Canada's national security and intelligence institutions — like the Canadian Security Intelligence Service, the Royal Canadian Mounted Police and the Canada Border Services Agency — continues to be bogged down by mistrust, says a new external report prepared for the federal government.
"We frequently heard about the trust gap between the country's national security institutions and Canadians, and in particular with racialized Canadians," says the report drafted by the National Security Transparency Advisory Group (NS-TAG) — an independent and external body first set up in 2019 to advise the deputy minister of Public Safety and the national security and intelligence community.
"At times, these relations have been marred by mistrust and suspicion, and by errors of judgment by these institutions, which impacted communities have perceived as discriminatory."
The NS-TAG group, made up of 10 members from legal, civil society and national security backgrounds, warns that the emergence of artificial intelligence and data-driven intelligence poses a threat to racialized communities.
"Systemic biases in Artificial Intelligence (AI) design can have perverse impacts on vulnerable individuals or groups of individuals, notably racialized communities," they found.
"These biases reflect not only specific flaws in AI programs and organizations using them, but also underlying societal cleavages and inequalities which are then reinforced and potentially deepened."
The report, published earlier this week, also calls on national security agencies to have better two-way conversations with communities.
"Too often engagement involves, in practice, government officials offloading a prepared message and failing to listen to the concerns of stakeholders," says the report.
"Constructive engagement should instead be based on dialogue; government officials should be attuned to the questions and concerns of stakeholders, listen to them, and be prepared and willing to respond."
The report also calls on agencies like CSIS to engage with communities on an ongoing basis — and not just when there's a crisis.
The authors pointed to CSIS's contact with the Iranian-Canadian community after the destruction of Flight PS752 in January 2020 and with the Muslim community following an attack on a mosque in Mississauga, Ont.
"Such engagement was important, but it was prompted by specific incidents. In our view, CSIS will not succeed in building long-term trust with racialized communities as long as its engagement is primarily reactive," says the report.
CSIS responded to the report's findings Friday by acknowledging the problem.
"We know that the voices of racialized communities and Indigenous peoples have not been heard as clearly as they should in conversations around policy, legislative and operational deliberations on national security matters," CSIS wrote in a response published Friday.
"We are committed to changing this."
The agency said it supports a number of the groups' recommendations and has promised to publish more details about its engagement programs in its annual reports.
The spy agency's director David Vigneault admitted in 2020 that CSIS has an internal racism problem as well.
"Yes, systemic racism does exist here, and yes there is a level of harassment and fear of reprisal within the organization," he said, according to a transcript of a 2020 meeting.
The agency said it's reviewing the report's section on artificial intelligence and has "taken note of the guiding principles and findings" in the report.
The other agencies have yet to issue a public response.