The federal cybersecurity centre says China, Russia and Iran are very likely to use tools enabled by artificial intelligence in attempts to meddle in the coming general election campaign.
In a new report, the Canadian Centre for Cyber Security says it expects individuals affiliated with the Chinese government will continue to target diaspora communities, pushing narratives favourable to Beijing's interests on social media platforms.
Cybercriminals are also likely to take advantage of election-related opportunities to perpetrate scams, says the centre, which is an arm of Canada's cyberspy agency, the Communications Security Establishment.
The centre says that despite these threats, it is very unlikely that AI-enabled activities will fundamentally undermine the integrity of the general election.
An election campaign is widely expected to begin shortly after the Liberals choose a new leader to succeed Prime Minister Justin Trudeau.
The federal centre says hostile foreign actors are using artificial intelligence to flood cyberspace with false information and are using botnets to spread this disinformation.
It warns that AI is also being used to create deepfake pornography targeting politicians and public figures — predominantly women and gender-diverse people.
An example of those tactics emerged Thursday.
Rapid Response Mechanism Canada, a federal unit that monitors the online environment for signs of foreign interference, detected a campaign to intimidate, belittle and harass people in Canada who criticize the Chinese government, said Global Affairs Canada.
Global Affairs said the campaign involved videos doctored through AI and the release of private information about targeted individuals without their consent — a tactic known as doxing.
The unit, known as RRM Canada, has been able to attribute the campaign, including the bots involved, to the Chinese government with high confidence, Global Affairs said in a statement.
The department described the campaign as "spamouflage" — a combination of spam and camouflage in which the content is distributed among everyday, human-interest style material.
"This is the first known instance where a Spamouflage campaign used sexually explicit deepfake photos to target an individual in Canada," said the statement.
The deepfake videos make fabricated claims about the prime minister, as well as other Canadian politicians and public figures.
The operation is a more extensive version of a campaign that Global Affairs reported on in 2023, the department added.
Global Affairs said it had contacted the victims of the new campaign, engaged with the relevant social media companies and raised concerns directly with the Chinese embassy in Canada.
