He didn’t say it, but the internet sure thinks he did.
A recent viral video of Canadian Prime Minister Mark Carney appearing to announce a ban on cars made before the year 2000 is a deepfake, cooked up by AI and convincing enough to fool almost anyone.
However, experts say there are signs people can look out for to separate the news from the fake news. Prof. Thomas Patrick Keenan from the University of Calgary says it all starts with the source.
“Did it come from a news outlet? Well, they probably fact checked it. Did it come from a TED Talk? Maybe not so good,” he says. “So I would say, start by considering the source.”
Keenan says if people are unsure, they should always check with a real clip and see if the voice doesn’t match.
“I think the way this one was done is that they used existing footage of Mark Carney saying something and put a totally different audio track on it,” he says.
To verify suspicious or AI-generated videos, Keenan recommends tools like InVID or submitting clips to fact-checking sites. But if you need a quick read, he says look closely at the visuals — blurry edges, mismatched lip-sync, or distorted logos can all be red flags.
The recent deepfake from Carney came from a TikTok account that his since been suspended.
CityNews took it to the streets to see if people could spot the deception. Everyone we spoke to thought it was a real clip.
It’s not the only deepfake of a political figure out there, and Keenan says we should only expect more as time goes on.
“These days, we are seeing more and more fake videos,” he says. “I mean, there’s famous ones of, you know, Justin Trudeau, and now it’s going to be Mark Carney.
“They go viral on the internet because they’re very entertaining, as long as you take them as entertainment and not as fact, that’s fine. They can be fun to watch. The tools to do this are getting so plentiful.”
