Facebook’s current stance on forged and fake content is pretty lenient. But, recently a very convincing deepfake video of Mark Zuckerberg hit the web, and now that leniency stands to change.
Hopefully, that is.
Earlier this year a similar fake video of Nancy Pelosi surfaced on the internet. In it, Pelosi is seen slurring her words, speaking as though she’d been hitting the bottle like a punching bag. The clip quickly went viral, (thanks in large part to Donald Trump and Rudy Giuliani) spreading like wildfire among conservative groups and individuals, across platforms and different social media outlets. It seemed as though Pelosi was in big trouble.
The video, of course, was faked. It had been slowed down and altered slightly to give the impression that Pelosi was sloshed during her television interview. It’s another one of the new “deepfake” videos that is blurring the lines between truth and lies, fact and fiction. As our technology gets better, so too does our ability to twist reality and spread mis-information. These deepfake videos are just the beginning of a strange and likely painfully confusing epoch of misrepresentation and outright lies.
Youtube removed the Pelosi video, saying that it violated its policies on “deceptive content.”
Facebook, however, did not. Instead they merely deprioritized the faked clip, tagged it as “false” and left it online for the world to see and share. Facebook’s director of public policy said, at the time, that if someone were to post a deepfake video of Mark Zuckerberg to Facebook, it, just like Pelosi’s, would have to stay up.
Not surprisingly, that bold statement was like a challenge to the deepfake geniuses at CannyAI. And not long afterwards we got this clip:
"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 deepfake Zuckerberg says in the video. "I owe it all to Spectre. Spectre showed me that whoever controls the data, controls the future."
Creepy stuff.
Now, while the voice actor’s Zuckerberg impression in the video isn’t exactly academy award winning acting, there is voice editing software out there that makes it totally possible to use someone’s actual voice, by splicing real audio clips of them from real content. Check out this Joe Rogan deepfake, which highlights the potential for this kind of voice forgery:
And, anyway, the deepfaked Zuck video is close enough to get the point across: unless we establish some very clear rules and regulations regarding deepfake content (and fast) no one is safe. Every public figure, every individual who has content online somewhere, is at risk of having their likeness hijacked and having their mouth stuffed full a stranger’s words.
Which is terrifying even before you start to consider some of the possibilities: deepfaked sex videos of politicians or celebrities, deepfaked clips of people inciting violence or saying hateful things, deepfaked videos of people fighting, abusing children, using drugs, or eating pizza with a knife and fork.
All of which will be so realistic, it will be very hard for people to tell the difference between fact and fiction.
That’s the situation we’re facing. And, it’s an insanely complex one. Because, on one hand, it seems very dangerous to put too much power of censorship into the hands of social media giants like Facebook and Twitter (just look at what happens to journalists who they disagree with).
On the other hand, if we don’t do something to regulate this kind of high-quality, high-resolution misinformation it could dissolve the very foundation of reality our society rests on. Which, is how we end up realizing an Orwell novel.
1984 ain't got shit on 2019.
Leave a Reply