In response to a growing problem on the Internet in which videos are being manipulated to distort reality — sometimes for political purposes — Facebook announced a new policy banning manipulated videos, commonly known as “deepfakes.”
MediaPost reports that Facebook Vice President of Global Policy Management Monika Bickert, who is scheduled to testify this week at a House hearing on “manipulation and deception in the digital age,” announced the new policy in a blog post on Facebook’s website.
“Manipulations can be made through simple technology like Photoshop or through sophisticated tools that use artificial intelligence or ‘deep learning’ techniques to create videos that distort reality — usually called ‘deepfakes,'” Bickert wrote. “While these videos are still rare on the internet, they present a significant challenge for our industry and society as their use increases.”
Bickert adds that Facebook will remove media that has been “edited or synthesized, beyond adjustments for clarity or quality, in ways that aren’t apparent to an average person and would likely mislead someone into thinking that a subject of the video said words that they did not actually say.” She also writes that the ban applies to content that is “the product of artificial intelligence or machine learning that merges, replaces or superimposes content onto a video, making it appear to be authentic.”
The policy does not apply to parodies or satire, MediaPost notes, adding: “Nor does it apply to video edited ‘solely to omit or change the order of words’ — although, if a Facebook fact checker rates a video as false or partly false, the platform will ‘significantly reduce its distribution in News Feed and reject it if it’s being run as an ad,’ says the post. Also, ‘people who see it, try to share it, or have already shared it, will see warnings alerting them that it’s false.’”
Videos of House Speaker Nancy Pelosi that have been edited to make her appear drunk are at the center of the controversy, but The Washington Post reports that the Pelosi videos would apparently not be banned by the new policy.
“Such a rule would still allow the infamously altered ‘drunk’ video of Pelosi that was viewed millions of times on Facebook last year, where her speech was slowed and distorted to make her sound inebriated,” The Post reports. “The effect was accomplished with relatively simple video-editing software. To contrast with more sophisticated computer-generated ‘deepfakes,’ disinformation researchers have referred to these kinds of videos as ‘cheapfakes’ or ‘shallowfakes.’”
Facebook’s Bickert adds in her post: “We are also engaged in the identification of manipulated content, of which deepfakes are the most challenging to detect. That’s why last September we launched the Deep Fake Detection Challenge, which has spurred people from all over the world to produce more research and open source tools to detect deepfakes. This project, supported by $10 million in grants, includes a cross-sector coalition of organizations including the Partnership on AI, Cornell Tech, the University of California Berkeley, MIT, WITNESS, Microsoft, the BBC and AWS, among several others in civil society and the technology, media and academic communities.”