AI-generated fake news and deepfakes spread in Canadian federal election campaign
AI-generated fake news stories and deepfake videos masquerading as legitimate Canadian media spread widely on social media platforms during the 2025 federal election campaign, including fraudulent investment schemes. Despite the volume of false content, researchers found no evidence that Canadian voters were manipulated or that election integrity was compromised.
What happened
AI-generated fake news stories masquerading as content from legitimate Canadian media outlets spread widely on social media during the federal election campaign. The content included fraudulent investment platforms purportedly endorsed by Liberal Leader Mark Carney featuring AI-generated images and videos, containing links to cryptocurrency scams.
AI-generated content also included memes and deepfake videos of politicians, as well as false news impersonating legitimate sources to promote fraudulent investment schemes involving cryptocurrency. About one-quarter of Canadians reported seeing such fake news stories, though a majority immediately recognized them as false.
Reported impact
Despite the prevalence of false information on social media, experts say the integrity of the election appears intact and Canadians have not been manipulated by the content. Researchers found more Canadians clicked and surfed websites of legitimate Canadian news organizations for election information this year than in previous elections.
Index analysis
The surge in AI-generated content filled an online news void created by the Online News Act, which required tech giants like Meta and Google to compensate journalism outlets for sharing their content. Meta removed news availability for Canadians on its platforms in August 2023, leaving the space where news would normally appear to be filled by unmoderated generative AI content.
The content became more sophisticated and politically polarizing as the campaign progressed. Foreign interference in the election campaign appeared to be 'relatively minor,' with some Chinese efforts to shape views about two politicians, though no evidence of successful voter manipulation has emerged.
Governance controls
Researchers noted that the AI-generated content was not properly moderated by Meta. The Media Ecosystem Observatory and the Canadian Digital Media Research Network tracked and analyzed the disinformation trends throughout the campaign.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.