A deepfake video of Mark Zuckerberg presents a new challenge for Facebook
Facebook refused to remove a doctored video of House Speaker Nancy Pelosi last month, but this week the company faces a new quandary: Will it remove a fake video of its own CEO, Mark Zuckerberg?
The video, posted to Facebook-owned Instagram over the weekend, falsely portrays Zuckerberg as saying, “Imagine this for a second: One man, with total control of billions of people’s stolen data, all their secrets, their lives, their futures.”
The video was made by taking 2017 footage of Zuckerberg and using artificial intelligence technology, known as deepfake technology, to manipulate Zuckerberg’s face to make it appear he said something he didn’t. Zuckerberg’s voice is replaced by an actor’s.
The video was created by an Israeli startup called Canny AI, cofounder Omer Ben-Ami confirmed to CNN Business on Tuesday. The video and several others — depicting celebrities like Kim Kardashian and US President Donald Trump — are part of a commissioned art installation called Spectre that was on display at the Sheffield Doc/Fest in the UK from June 6-11. The videos, which Canny and a couple other companies worked with artists to make, are meant to show how technology can be used to manipulate data.
Last month Facebook declined to remove a manipulated video that made it appear House Speaker Nancy Pelosi was slurring her words. The company said it downranked the false video, meaning it would be seen in fewer people’s Facebook News Feeds. Pelosi later blasted the company, “I think they have proven — by not taking down something they know is false — that they were willing enablers of the Russian interference in our election,” she told California’s KQED.
Asked whether a doctored video of Zuckerberg would get the same treatment as the manipulated Pelosi video, Neil Potts, Facebook’s director of public policy, told a parliamentary hearing in Canada last month, “If it was the same video, inserting Mr. Zuckerberg for Speaker Pelosi, it would get the same treatment.”
The Pelosi video was not a deepfake like the new Zuckerberg clip, but was manipulated using traditional video editing techniques. CNN Business has asked Facebook if it has any specific rules for deepfakes.
The Zuckerberg video, which was first reported by Vice, comes as the US Congress prepares to hold its first hearing on the potential threats posed by deepfake videos. Earlier this year, the US Director of National Intelligence warned that America’s adversaries may use deepfake technology in future disinformation campaigns targeting the country.
Until recently, video hoaxes were relatively rare since they are harder to pull off than fakes of still images, but this is changing rapidly thanks to the rise of GANs, or generative adversarial networks. GANs can use data to produce new things. The technique is also used for making deepfakes.
In this case, Ben-Ami said, Canny AI chose a scene of less than a minute in length of Zuckerberg speaking, and used a computer to meld it with an actor’s voice and appropriate facial movements. It took about a day to make an initial version of the video with one of the artists’ voices standing in for a voice actor’s Zuckerberg impression and facial movements, he said. It took another two to three hours to make the final version.
CNN Business reached out to the artists behind Spectre but did not immediately receive a response.
Ben-Ami, whose company focuses on dubbing speech in videos from one language to another, is concerned about the video and others Canny AI made spreading without the context that they were created as art. But he said he also wants to raise questions about the creation of such media.
“People need to know it’s possible to do it,” he said.
An Instagram spokesperson told CNN Business that the site will treat the video “the same way we treat all misinformation on Instagram.” If it’s marked as false by third-party fact checkers, the spokesperson said, the site’s algorithms won’t recommend people view it.
Last month, Facebook announced it would spend millions to fund academic institutions studying image and video analysis. Some of that work involves tackling deepfakes, the company said.