MENLO PARK, CALIFORNIA — Facebook has put forth a new policy that would ban all videos that are "edited or synthesized" using AI technology in a way that would be difficult for average Facebook users to spot, reports the Washington Post.
This would include a ban on videos that use techniques to make a subject say words that he or she never said.
Though the company said in a news release that they will allow manipulated videos as long as they are for satire or parody purposes.
In this case, the Washington Post points out that under the policy, a doctored video showing Speaker of the House Nancy Pelosi drunkenly giving a speech last year would still be allowed on Facebook.
Last year, the company declined to remove the video even though fact-checkers had determined that the video was fake.
A Facebook spokesperson explained to the Washington Post that this is because the company doesn't have a policy that says that the information you post on their platform must be true.
Facebook's new policy would also not restrict simpler forms of video deception.
This includes mislabeled footage, or trimming a video to make something a person says sound out of context.
Another Facebook spokesperson said in a statement that the company would remove political ads if they include highly manipulated videos.
A digital forensics expert from UC Berkeley told the Washington Post in an email that the social media giant should focus on the issue of removing misleading videos in general instead of just deepfake videos.