Canadian Researchers Develop AI Tool to Combat Online Disinformation

Canadian researchers are turning to artificial intelligence to combat a growing wave of online disinformation targeting national unity and public perception, according to experts leading the effort.

The Canadian Institute for Advanced Research has enhanced its disinformation detection tool, CIPHER, with AI capabilities to keep pace with the increasing volume of false and misleading claims circulating online. The system, which initially focused on Russian disinformation campaigns, is now expanding to analyze Chinese-language content and potentially material originating from the United States.

“Russia was the main threat targeting Canada most generally,” said Brian McQuinn, an associate professor at the University of Regina and co-lead of the CIPHER project. “We are now beginning to shift.”

McQuinn cited a recent example where the system identified a Russian media outlet falsely reporting that Alberta is moving toward independence. While separatist movements exist in the province and have reportedly engaged with U.S. officials, no formal separation process is currently underway.

“Effective disinformation often has kernels of truth in it,” McQuinn explained.

CIPHER was launched three years ago following a report by McQuinn and colleagues that uncovered pro-Kremlin social media accounts targeting both far-right and far-left groups in Canada with false narratives about the war in Ukraine. These included unfounded claims that Russia invaded to eliminate a neo-Nazi regime and that Ukraine had pursued nuclear weapons.

According to McQuinn, the ultimate goal of these disinformation campaigns is to fracture social cohesion and potentially provoke violence. The campaigns become particularly effective when ordinary citizens share the content with their personal networks.

“It is essential for China and for Russia, especially, to show that it looks like the Western project is decaying, is falling apart economically, politically, socially,” he said.

The researchers have identified the United States as an increasingly significant source of disinformation affecting Canada, noting that most Canadian social media discourse takes place on American platforms.

“We have seen that Canadian news and certain types of Canadian content are being downgraded and throttled within these algorithms,” McQuinn observed.

While artificial intelligence has contributed to the proliferation of disinformation on social media, the CIPHER team recognized that the same technology was necessary to streamline their fact-checking efforts. “We are in an AI arms race around disinformation,” McQuinn acknowledged.

The researchers aim to make CIPHER available to government agencies and non-profit organizations. Currently, the tool is being utilized by DisinfoWatch, an organization dedicated to exposing falsehoods to the Canadian public.

Marcus Kolga, founder of DisinfoWatch, has called for stronger legislation and regulations on digital platforms to prevent the spread of misinformation through social media accounts.

“Us doing it alone is not sufficient enough. It requires technology and for us to harness existing technologies in order to sort of make up that gap that we have,” Kolga stated.

McQuinn revealed that discussions have taken place with government agencies regarding the potential adoption of CIPHER, though he declined to provide specific details. The Canadian Institute for Advanced Research has received funding support from both the federal government and the province of Alberta.

To combat the spread of disinformation, McQuinn advises Canadians to pause before sharing content on social media. “If I’m going to forward something, what am I forwarding?” he suggested. “The research has shown if you just take like an extra 10 seconds, the amount of disinformation that gets transferred is significantly less.”

As foreign influence operations continue to evolve, tools like CIPHER represent a critical counterbalance in preserving information integrity within Canada’s digital landscape. The intersection of human fact-checking expertise and advanced AI capabilities offers a promising approach to identifying and countering sophisticated disinformation campaigns targeting Canadian society.