China, Russia and Iran could try to use artificial intelligence tools to interfere in Canada’s general election this year.
Canada is particularly vulnerable as it “increases exposure to AI-enabled malignant impact campaigns” as the majority of its citizens receive news and information from the Internet or social media. Meanwhile, not only data from Canadians, but public and political organizations can also be mined from online sources, allowing foreign actors to create fake content and create tailored propaganda campaigns.
“The PRC, Russia and Iran rated that they are very likely to use AI-enabled tools to try to hamper Canada’s democratic processes before and after the 2025 election,” the report reads.
“When targeting Canadian elections, threat actors are most likely to use generator AI as a means of creating and disseminating disinformation designed to sow division among Canadians and drive narratives that encourage foreign interests.”
The use of generated AI by hostile actors interfering with elections around the world, including Europe, Asia and the US has been increasing over the past two years, with 102 cases of these tools being reported to interfere or influence the 41 elections held between 2023 and 2024, the report states. Note that generation AI tools can create new content based on specific data, such as new “text, images, audio, video, or software code.”
China’s threat to Canadian elections
The People’s Republic of China (PRC) is the most likely foreign actor to target Canadian elections, saying that Russia and Iran view Canadian elections as “almost certainly” a lower priority target compared to US and UK elections. If Russia or Iran is targeting Canada, “they are more likely to use low-effort cyber or impact operations.”
Related Stories
The institution cites the 2021 Canadian General Election as an example. In the Canadian general election, actors affiliated with the Chinese administration spread non-AAI disinformation about politicians running for positions considered “anti-PRC.”
Two years later, the propaganda campaign, called the “Spamouflage Dragon,” which is likely linked to China, targets dozens of lawmakers, including Prime Minister Justin Trudeau, conservative leader Pierre Polyeille, and several cabinet members, says the network used previously used AI to target Canadian Mandarin language.
Another important finding in the report is that some foreign countries, particularly the PRC, are running a “large-scale data collection campaign” targeting “democratic politicians, public figures and citizens around the world.”
Advances in predictive AI will increase risk as these states “quickly queri and analyze these data.” This will help to deepen our understanding of the political environment of democratic nations, the report states. Predictive AI tools are designed to analyse data by recognizing patterns in the data instead of generating new content.
“By owning a detailed profile of key targets, social networks and voter psychology, threat actors almost certainly strengthen their targeted influence and ability to carry out spy campaigns,” the report states.
The report points out that the Chinese administration “probably” used social media platform Tiktok to promote the narrative of the democratic state Pro-PRC, which it identifies as anti-PRC. The work aimed at “user beliefs and behaviors affect behavior on a large scale” is “highly likely” to target voters ahead of at least one election. Tiktok is owned by PRC-based company Bytedance.
“We appreciate the very high chance that actors associated with the PRC will continue to specifically target Canadian Chinese and diaspora communities, and are pushing for a narrative that favors PRC’s interest on social media platforms,” the report said.
Election integrity
While it is “highly unlikely” that disinformation or AI-enabled cyber activity will undermine the integrity of Canada’s upcoming general elections, the advancements in AI and the increased capabilities of cyber enemies in using these technologies “are likely to increase the threat to future Canadian general elections,” CSE said in its report.
Canada holds general elections by paper votes, but much of its election infrastructure is digitalized, including voter registration systems, election websites and communications within election agencies, making these systems vulnerable to malicious cyber activity.
“Cyberactors can quickly create targeted, persuasive phishing emails using generative AI. This could allow illegal entries into this infrastructure, which can install malware and exclude sensitive information and publish it,” the agency says.
The report cites a lawsuit from last July that allows hackers from China to access UK election registration registrations in the names and addresses of all people registered to vote between 2014 and 2021, according to the UK government. “AI-enabled cyberactors can use this kind of data to develop propaganda campaigns tailored to a specific audience,” the CSE report said.
The agency says it suspects foreigners will “enforce destructive cyberattacks on election infrastructure, including attempting to paralyze telecommunications systems on Election Day.” Meanwhile, Canadian politicians and political parties point out that they are likely targets of “hack and leak operations” that involve stealing victims.‘ We publish information and sensitive data for purposes such as seeking financial benefits, reputational damage or crippling organizations.
