Facebook Tested With Deepfake Of Mark Zuckerberg: Company Leaves It Up

(Mis)Uses of Technology

from the as-it-should dept

Over the last few weeks there’s been a silly debate over whether or not Facebook made the right call in agreeing to leave up some manipulated videos of House Speaker Nancy Pelosi that were slowed down and/or edited, to make it appear like she was either confused or something less than sober. Some Pelosi-haters tried to push the video as an attack on Pelosi. Facebook (relatively quickly) recognized that the video was manipulated, and stopped it from being more widely promoted via its algorithm — and also added some “warning” text for anyone who tried to share it. However, many were disappointed that Facebook didn’t remove the video entirely, arguing that Facebook was enabling propaganda. Pelosi herself attacked Facebook’s decision, and (ridiculously) called the company a “willing enabler” of foreign election meddling. However, there were strong arguments that Facebook did the right thing. Also, it seems worth noting that Fox News played one of the same video clips (without any disclaimer) and somehow Pelosi and others didn’t seem to think it deserved the same level of criticism as Facebook.

Either way, Facebook defended its decision and even noted that it would do the same with a manipulated video of Mark Zuckerberg. It didn’t take long to put that to the test, as some artists and an advertising agency created a deep fake of Zuckerberg saying a bunch of stuff about controlling everyone’s data and secrets and whatnot, and posted it to Facebook-owned Instagram.

And… spoiler alert: Facebook left it up.

?We will treat this content the same way we treat all misinformation on Instagram,” a spokesperson for Instagram told Motherboard. “If third-party fact-checkers mark it as false, we will filter it from Instagram?s recommendation surfaces like Explore and hashtag pages.?

This actually is not a surprise (nor should it be). People keep wanting to infer nefarious intent in various content moderation choices, and we keep explaining why that’s almost never the real reason. Mistakes are made constantly, and some of those mistakes look bad. But these companies do have policies in place that they try to follow. Sometimes they’re more difficult to follow than other times, and they often involve a lot of judgment calls. But in cases like the Pelosi and Zuckerberg manipulated videos, the policies seem fairly clear: pull them from the automated algorithmic boost, and maybe flag them as misinformation, but allow the content to remain on the site.

So, once again, we end up with a “gotcha” story that isn’t.

Of course, now that Pelosi and Zuck have faced the same treatment, perhaps Pelosi could get around to returning Zuckerberg’s phone call. Or would that destroy the false narrative that Pelosi and her supporters have cooked up around this story?

Oh, and later on Tuesday, CBS decided to throw a bit of a wrench into this story. You see, the fake Zuckerberg footage is made to look as though it’s Zuck appearing on CBS News, and the company demanded the video be taken down as a violation of its trademark:

Perhaps complicating the situation for Facebook and Instagram a call late Tuesday from CBS for the company to remove the video. The clip of Zuckberg used to make the deepfake was taken from an online CBS News broadcast. “CBS has requested that Facebook take down this fake, unauthorized use of the CBSN trademark,” a CBS spokesperson told CNN Business.

Of course, if Facebook gives in to CBS over this request, it will inevitably (stupidly) be used by some to argue that Facebook used a different standard for disinformation about its own exec, when the reality would just be a very different kind of claim (trademark infringement, rather than just propaganda). Hopefully, Facebook doesn’t cave to CBS and points out to the company the rather obvious fair use arguments for why this is not infringing.

Filed Under: content moderation, deep fake, deepfake, mark zuckerberg, nancy pelosi
Companies: facebook, instagram