Day 6

What a deepfake video of Mark Zuckerberg reveals about how we're manipulated online

This week, a deepfake video of Facebook CEO Mark Zuckerberg went viral. Many viewed it as a strategic takedown of the company's fake news policies, but video co-creator Bill Posters says it was actually part of something much bigger.

The video was actually part of an art installation critiquing data privacy

Mark Zuckerberg, chief executive officer and founder of Facebook Inc., listens during a House Energy and Commerce Committee hearing in Washington, D.C., on April 11, 2018. (Andrew Harrer/Bloomberg)

Bill Posters set out to critique the manipulation of social media users, and ended up manipulating social media users.

The Sheffield, U.K., artist posted a deepfake video of Facebook founder Mark Zuckerberg to Instagram last week, complete with CBS News branding.

In the video, Zuckerberg purportedly says, "One man with total control of billions of people's stolen data … controls the future."

The video is doctored — created with sophisticated algorithms designed to simulate natural facial movements. But unlike most deepfakes, Posters doesn't try to hide the fact it's artificial: the video includes hashtags like #deepfake and #digitalart.

The video is actually a fake endorsement for a made-up tech company called Spectre, and part of an art installation aimed at critiquing the power tech giants have over what their users see.

"We always set out to position the Spectre installation as a kind of critique of the digital influence industry," Posters told Day 6 host Brent Bambury.

"Now we're finding ourselves in the middle of the teeth of the industry because of all the viral activity online."

Unregulated, non-transparent tech

The fake Zuckerberg video comes just weeks after a doctored video of U.S. House Speaker Nancy Pelosi, apparently slowed down to make her speech seem slurred, was posted to Facebook.

Despite the company's own fact-checkers determining the video is fake, Facebook refused to remove the video of Pelosi, saying it would instead be removed from searches and accompanied by fact-checked articles.

On Tuesday, Facebook-owned Instagram said they would follow the same principles, opting to keep the Zuckerberg deepfake available on the photo-sharing platform.

When Posters and his collaborator Daniel Howe created Spectre alongside AI researchers that produce deepfake videos, part of project involved exploring how they would be treated by the platforms.

"All of these technologies — whether they're deepfake technologies or micro-targeted behavioural profiling technologies that exist within the digital influence industry — they're all operating in an unregulated, totally non-transparent environment right now," Posters said.

Zuckerberg isn't the only victim of Posters' deepfake experiment: U.S. President Donald Trump and reality television star Kim Kardashian are among his company.

By refusing to remove the videos, Posters said Facebook arguably proved his point.

'Visitors can feel what's at stake'

As a physical installation, Spectre just wrapped a run at the Sheffield Documentary Festival in the U.K.

When visitors walk in, they're greeted by a circle of six two-metre tall black monoliths equipped with touch screens, representing the black box servers that power the internet.

It's a scene reminiscent of Stanley Kubrick's 2001: A Space Odyssey.

Spectre is an interactive art installation that offers users a look inside how their data is used, but only if they're willing to 'pray at the altar of Dataism'. (Submitted by Bill Posters)

"When you log in with some basic details, Spectre tries to indoctrinate you into this new religious philosophy that is 'dataism' by showing you how powerful its products are," Posters told Bambury.

Spectre then attempts to sell customers what they desire by offering four options — truth, power, influence or wealth. Users then take part in a 30-second game where they rank the brands they love, and the brands they hate.

Spectre takes that data and compares it with more than a million other personality profiles, provided by the University of Cambridge Psychometric Centre, in order to create a personalized user profile.

"We use the same tactics that the industry uses, which are kind of like obfuscation tactics where data extraction is hidden by behavioural design or by gamification," Posters said.

Based on their new profile, users are shown how their data is used to influence their decisions not only online, but in shopping malls and during elections. 

"Visitors can feel what's at stake when our data is used in really opaque ways by powerful tech giants," Posters added.

'Positive' impact

According to Posters, the installation has had a profound and emotional effect on its viewers, many of whom are accustomed to ignoring the detailed legalese of every new usage policy they're presented.

"We always say 'yes' to the permissions to view articles or to get access to maps or to download content," he said.

"We would hope that people will just have a deeper engagement and understanding about what it feels to have our data used in expected ways and how that then operates in the real world."

When it comes to the deepfake videos, Posters says he's happy that they've sparked a conversation about Silicon Valley's influence over its customers.

"So many of the technologies that are used to kind of extract our data and to influence our behaviours are just so obscure and hard to understand," he said.

"We're just very grateful that we can use these forms of digital to engage audiences in a really accessible and relational way."


To hear more from Bill Posters, download our podcast or click 'Listen' above.