Welcome to our new series where we fact-check deepfakes that are circulating on the web. Read on for a glimpse of how damaging, disruptive, and deceptive AI-manipulated content can be.
The Mark Carney Deepfake refers to a viral 28-second video from May 2025 that falsely showed Canadian Prime Minister Mark Carney announcing a ban on vehicles made before 2000. Fact-checking confirmed the clip was an AI-generated deepfake created from altered footage of a real March 27, 2025 press conference, with clear facial, audio, and visual anomalies. The incident highlights the growing use of deepfakes to spread political misinformation during sensitive periods.
Case File
In early May 2025, a video surfaced online purporting to show Canadian Prime Minister Mark Carney announcing a ban on vehicles manufactured before the year 2000. This video quickly went viral across social media platforms, sparking public confusion and concern. However, upon closer examination, it became evident that the video was a deepfake – a manipulated piece of media designed to mislead viewers.
Deepfake Detector Analysis
Video Length: 28 seconds
Detection Time: Under 2 seconds
Number of Alerts: 12
Anomalies Detected:
- Facial Inconsistencies: Subtle glitches in lip synchronization and unnatural eye movements.
- Audio Discrepancies: Mismatched vocal tone and cadence compared to verified recordings of PM Carney.
- Visual Artifacts: Distortions around the mouth and jawline during speech.
The Reality
Investigations revealed that the deepfake was crafted using footage from a legitimate press conference held by PM Carney on March 27, 2025. In the original event, he discussed Canada’s response to U.S. tariffs and mentioned plans to procure Canadian-made vehicles for federal use. There was no mention of banning older vehicles. The deepfake manipulated this footage to fabricate a policy announcement that never occurred.
Broader Implications
This incident is part of a troubling trend where deepfakes are used to spread misinformation, particularly during politically sensitive times. A report by Canada’s Media Ecosystem Observatory highlighted a surge in fake political content on social media leading up to the federal election, with over a quarter of Canadians encountering such disinformation online.
For an idea of how the public responded to this deepfake video, just take a look at some of the responses to this post on X.
Get Ahead of the Problem
At X-PHY, we recognize the dangers posed by deepfakes and are dedicated to combating them in an easy, lightweight, and effective manner.
The X-PHY Deepfake Detector delivers this through on-device processing that ensures your safety and privacy. Contact us to learn more or visit our eStore today. Special pricing for enterprise deployment is available upon request.
