Social media giant Facebook has announced that it won’t take down a doctored video – a so called ‘deepfake’ – of the company’s CEO Mark Zuckerberg from Instagram, where it was posted last week.
As an Instagram representative told Motherboard, they intend to treat the video, created by artists Bill Posters and Daniel Howe in partnership with the advertising company Canny, the same way they treat “all misinformation” posted on the platform.
"If third-party fact-checkers mark it as false, we will filter it from Instagram’s recommendation surfaces like Explore and hashtag pages", they said.
Посмотреть эту публикацию в Instagram
The video, which is based on the actual September 2017 address by Zuckerberg regarding the alleged Russian interference in the American election via Facebook, features his likeness saying the following: "Imagine this for a second: One man, with total control of billions of people's stolen data, all their secrets, their lives, their futures. I owe it all to Spectre. Spectre showed me that whoever controls the data, controls the future."
As the media outlet points out, this development occurred after Facebook “has been forced to take a stance on whether fake or altered images are allowed to stay up on the site”, when a similarly manipulated video of House Speaker Nancy Pelosi went viral.
In response, Facebook’s director of public policy Neil Potts confirmed that if someone were to post a doctored video of Zuckerberg, it would stay up, just like the video of Pelosi.
Pentus-Rosimannus: If instead of Pelosi, a doctored video of Mark Zuckerberg was slowed down to make him seem impaired, would you leave it up?
— Alex Boutilier (@alexboutilier) 28 мая 2019 г.
Potts: Yes.
The “Zuckerberg deepfake” wasn’t the only such video made by Canny in collaboration with Posters, with manipulated videos of US President Donald Trump and reality TV star Kim Kardashian also being available online.
Посмотреть эту публикацию в Instagram
Посмотреть эту публикацию в Instagram
As one of Canny’s founders, Omer Ben-Ami, explained, the company “used a proprietary AI algorithm, trained on 20 to 45 second scenes of the target face for between 12-24 hours” in order to make these fake videos.
