How will Facebook handle AI-generated fake videos? Well, don't expect the company to take them down.
On Tuesday, Facebook said it would leave up an AI-generated video falsely depicting company CEO Mark Zuckerberg talking about an Illuminati-like secret society out to control the world.
"Imagine this for second: One man with total control of billions of people's stolen data," the fake Zuckerberg says in the video. "All their secrets. Their lives. Their futures. I owe it all to Spectre. Spectre showed me that whoever controls the data, controls the future."
The video looks realistic, but it's actually a product of an AI-powered algorithm; the software was able to change Zuckerberg's lips from a 2017 interview to synch with the phony dialogue spoken over the video.
Most people can probably tell the footage is fake simply by listening to voice, which doesn't fully match how the real Zuckerberg speaks. Nevertheless, the technology behind the video has had IT experts and lawmakers worried. Could it one day be used to fool the public?
The "deepfake" video, which was first reported by Motherboard, has had journalists wondering whether the social network would delete it after the doctored footage was uploaded to the Facebook-owned Instagram. The answer is no. Instead, the company might decide to limit its circulation.
"We will treat this content the same way we treat all misinformation on Instagram. If third-party fact-checkers mark it as false, we will filter it from Instagram's recommendation surfaces like Explore and hashtag pages," an Instagram spokesperson said in a statement.
Although Facebook has been trying to fight false information over the platform, the company often resorts to suppressing the content rather than deleting it. That's because the social network has been trying to maintain a balance between abuse and free speech, including comments that can be extreme or offensive.
In this case, the doctored video of Zuckerberg comes from a pair of artists who partnered with a technology startup called Canny AI to create the footage. Their project, dubbed Spectre, seeks to highlight how social media can be used to manipulate the public. To illustrate this, the artists have created other deep fakes depicting President Donald Trump and reality TV star Kim Kardashian in a nefarious light.
"I feel really blessed because I genuinely love the process of manipulating people online for money," says Kardashian in the deep fake video.
Still, Facebook is facing pressure to crack down on other doctored videos. Last month, the company's attempts to maintain a balance between abuse and free speech were challenged when edited clips of House Speaker Nancy Pelosi depicting her slurring her words spread over the platform. In response, the company refrained from deleting the content. Instead, it began trying to limit their circulation and using a disclaimer to warn that the videos were false.
Whether this approach will work in the event Facebook is flooded with deepfake videos remains to be seen. Others have complained the company is risking political censorship with content regulation. In the meantime, a Congressional committee plans on holding a hearing this Thursday examining the national security dangers of AI-generated videos.