Deepfake Network Used Canadian Creators’ Videos to Push Political Claims to 22 Million Viewers

November 13, 2026
Deepfake Network Used Canadian Creators’ Videos to Push Political Claims to 22 Million Viewers
  • A McGill University report found a likely foreign-based network using deepfakes of real influencers to spread politically divisive falsehoods in Canada, combining familiar faces with fabricated lip movements and AI-generated voices.

  • Researchers identified 293 TikTok and Instagram accounts that posted more than 3,200 videos over three months, generating 22.1 million views; the content touched on Alberta’s referendum, provincial elections and U.S.-Canada trade tensions.

  • The creators whose likenesses or content were used included Kristina Collins, known as Kallmekris, conservative commentator Jasmin Laine and mortgage broker Ron Butler, alongside creators with a range of perspectives.

  • Trade-war videos carried misleading claims from both Canadian and American perspectives, including a fabricated story that Donald Trump gave Canada 48 hours to remove retaliatory tariffs and that Mark Carney then cancelled $8-billion in U.S. defence contracts; 161 videos using the script drew about 5.1 million views.

  • Some clips repurposed creators’ footage with new backgrounds, while others used AI-generated faces, voices, lip movements and scripts to make people appear to say things they never said, including Edmonton commentator Cole Bennett’s referendum message.

  • The videos varied in accuracy and political stance, and Canadian users interacted with them without realizing they were fake; Laine said fabricated clips prompted accusations that she was paid to repeat claims she had never made.

  • The report found that only 6.3 per cent of commenters identified the videos as fake or propaganda, while some addressed the narrators as if they were real people.

  • TikTok’s AI-labeling system flagged only 4 per cent of the videos, and researchers suggested that incorporating real creators’ footage may help altered clips evade detection.

  • The organizers and their purpose remain unknown, but account registrations and posting patterns point to activity outside North America, possibly in China; researchers found no clear evidence of monetization or a consistent foreign-government narrative, though they said the apparent aim was to polarize audiences.

  • Researchers said the operation could be building audiences for later use, testing a tool or service, or acting for someone seeking to influence Canadians.

  • After the report was released on November 12, TikTok said it permanently banned the identified accounts and removed additional accounts and content for violations involving deceptive behaviour and misleading AI-generated material.

  • TikTok said its moderation technology detects or removes more than 98 per cent of deceptive-behaviour content before users report it; Meta declined to comment, and an Instagram spokesperson did not immediately say whether the videos had been removed.

Summary based on 9 sources


Get a daily email with more Tech stories

More Stories