As artificial intelligence advances, the capacity for malicious actors to covertly spread misinformation has also increased. To mitigate this, Facebook has announced that the company will no longer allow “deepfake videos” or “misleading manipulated media” on its platform.
But in order for content to be banned, it must meet a strict criteria. First, it must have “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 they did not actually say,” writes Monica Bickert, the vice president for global policy engagement at Facebook, in a blog post published on Monday.
In addition, the content must be “the product of artificial intelligence or machine learning that merges, replaces or superimposes content onto a video, making it appear to be authentic.” The blog post also notes that the policy “does not extend to content that is parody or satire, or video that has been edited solely to omit or change the order of words.”
According to The New York Times, “deepfakes” that meet both of the above criteria will also be banned from political ads.
After Facebook announced the new policy, left-wing outlets were quick to imply that the policy was not comprehensive enough. While Vox lamented that “controversial fake videos like the one of Nancy Pelosi that made her appear to be drunk” would not be banned, The Washington Post, which has called the Pelosi video “doctored,” noted in a headline about the policy change that the “new policy may not cover the controversial Pelosi video.”
Earlier this week, the U.S. House Committee on Energy & Commerce issued a memo that described “deepfakes” as “videos that have been manipulated using machine learning technology such as a Generative Adversarial Network, or GAN, to look and sound real. These videos can show people doing and saying things they never did or said. The GAN process detects flaws in the fake video, which is then used to improve the fake video.”
The Nancy Pelosi video does not seem to meet this threshold.
Facebook has also been subject to criticism for refusing to remove factually incorrect (in the eyes of at least some “fact-checkers”) content from its platform, but the tech giant has repeatedly defended its decision that the company should not attempt to serve the role of arbiter of truth.
“If we simply removed all manipulated videos flagged by fact-checkers as false, the videos would still be available elsewhere on the internet or social media ecosystem,” writes Bickert in the blog post. “By leaving them up and labelling them as false, we’re providing people with important information and context.”
“Our enforcement strategy against misleading manipulated media also benefits from our efforts to root out the people behind these efforts,” continues Bickert. “Just last month, we identified and removed a network using AI-generated photos to conceal their fake accounts. Our teams continue to proactively hunt for fake accounts and other coordinated inauthentic behavior.”
The Facebook announcement comes on the heels of a House subcommittee meeting being held today about deception in the digital age and the risks associated with “deepfake” technologies. Bickert is expected to testify.