China, Russia will ‘very likely’ use AI to target Canadian voters: Intelligence agency
Canada’s cyber intelligence agency, the Communications Security Establishment (CSE), has issued a warning that countries such as China, Russia, and Iran are likely to use artificial intelligence (AI) to interfere in the upcoming federal election. The CSE report highlights the use of AI to fuel disinformation campaigns and hacking operations, posing a threat to Canada’s democratic process.
According to the 28-page document released by the CSE, hostile actors are expected to leverage generative AI to create and spread misinformation, sow division among Canadians, and promote narratives that align with foreign interests. This technology, which powers tools like ChatGPT, can generate text, images, music, and video by analyzing vast amounts of data. The report emphasizes the risk of AI-enabled activities in undermining the integrity of the election but states that it is “very unlikely” to fundamentally disrupt the electoral process.
Bridget Walshe, the associate head of the CSE’s Canadian Centre for Cyber Security, highlighted the potential of AI to automate the creation and dissemination of misinformation on social media platforms at an industrial scale. She noted that threat actors could use generative AI to engage in phishing attempts and hack-and-leak operations targeting Canadian politicians and political parties.
Recent incidents of foreign interference, such as the disinformation campaign against Liberal leadership candidate Chrystia Freeland, have raised concerns about the vulnerability of Canadian political figures to cyber threats. The CSE report warns that foreign actors could exploit the vast amount of data held by Canadian political parties to launch targeted influence campaigns and espionage operations.
While the CSE believes that hostile actors are unlikely to carry out destructive cyberattacks on election infrastructure, it cautions that the use of AI for disinformation campaigns remains a significant threat. The report identifies China as the most likely actor to deploy AI capabilities to influence Canadian voters, particularly within Chinese-diaspora communities. Russia and Iran are also mentioned as potential threats, albeit with lower priority targets compared to the US and UK.
Despite the risks posed by AI-enabled disinformation, the CSE and its cyber center are prepared to respond to potential threats to the upcoming election. The agency has provided guidance to political parties and candidates and collaborates with Elections Canada to ensure cybersecurity during the election process.
As Canada braces for a potential election in the spring, the CSE’s warning underscores the need for vigilance against AI-driven disinformation campaigns. With the threat landscape evolving rapidly, it is crucial for political parties and stakeholders to enhance their cybersecurity measures and safeguard the integrity of the democratic process.
