Facebook to issue rules banning deepfakes: report

Facebook will issue new rules as early as Tuesday banning users from posting so-called deepfakes, The Washington Post reported.

The regulations would prohibit users from publishing computer-generated, manipulated videos ahead of the 2020 election in an effort to limit disinformation, three people familiar with the matter told the Post. 

The rules would not forbid all edited videos, only those that are “edited or synthesized” by technologies like artificial intelligence that the average user would not be able to catch, the sources told the newspaper.

This would not include videos that were edited for satire or parody purposes or those with minor manipulation like mislabeling footage, interweaving dialogue and taking quotes out of context. The social media outlet could fact-check and regulate the spread of videos altered in more minor ways.

Under these rules, the clip that went viral last year that edited Speaker Nancy Pelosi’s (D-Calif.) speech to make her sound drunk would not have been prohibited, the Post reported.

The regulations uses similar language discussed in a June 2019 meeting in San Francisco called by the Carnegie Endowment for International Peace. A person present at the meeting told the newspaper there was debate over what level of editing is required to define a video as misleading and whether social media networks need stricter rules ahead of 2020.

The Hill reached out to Facebook for comment.