Skip to main content
RBC
Deepfake technology is rapidly improving—and we can't look away.

The hottest app in China right now, Zao, swaps your face with Leonardo DiCaprio’s to amusing effect. Suddenly, you’re Jack on the bow of the Titanic. It’s fun, silly—and a peak at a looming challenge to our democracy.

Right now, it’s still relatively easy to spot a deepfake with the naked eye. The voice isn’t quite right, or the blinking pattern is unnatural. But as the machine learning that powers these impersonations gets better and better, the joke will be on us. In just a few short years, we won’t be able to believe our eyes and ears anymore.

Deepfakes pose a big threat to public trust, misleading people and spreading false information. When you consider 34% of Americans say the Internet is their preferred format for news consumption, the potential impact of deepfakes on election campaigns is staggering.

Already, ahead of the 2020 U.S. election, a video of House Speaker Nancy Pelosi slurring her way through a speech was viewed millions of times online, tricking people into thinking she was drunk. It took several days to clarify what had happened: the viral video wasn’t real; it had been doctored.

The truth is in danger. That’s why, with the Canadian federal election underway, we hosted an RBC Disruptors conversation about how technology is transforming the vote. You can watch the full session, which featured panelists:

  • Zeynep Tufekci, Techno-Sociologist & Associate Professor, University of North Carolina
  • Kevin Chan, Global Director and Head of Public Policy, Canada, Facebook
  • Shuman Ghosemajumder, Chief Technology Officer, Shape Security

Now that we are waking up to the real threat posed by deepfakes, who is responsible for fighting back? And how? We don’t know how many fake videos are already in circulation, but we do know it doesn’t take long to create one: Shape Security wrote a blog detailing how they turned me into Simon Cowell as a fun side project in just a few days.

In our latest RBC Disruptors podcast episode, Shape Security’s Shuman Ghosemajumder explains the making and propagating of deepfakes, and what legislators, Big Tech and ordinary citizens can do to protect our democratic institutions.