WASHINGTON, UNITED STATES –
Facebook will ban in-depth videos before the US elections, but the new policy will still allow much edited clips, provided they are parody or satire, the social media giant said Tuesday.
Deepfake videos are hyperrealistic manipulated clips made using artificial intelligence or programs that are designed to accurately falsify real human movements.
In a blog post following a Washington Post report, Facebook said it would begin to remove edited clips – in addition to clarity and quality – in ways that are "not apparent to the average person" and may mislead people.
The clips would be removed if they were "the artificial intelligence or machine learning product that blends, replaces, or superimposes content on a video, making it look authentic," said Facebook Vice President Monika Bickert.
However, the statement added, "This policy does not extend to content that is parody or satire, or video that was edited solely to omit or change the word order."
The US media noted that the new guidelines would not cover videos such as the 2019 viral clip – which was not a failure – by House Speaker Nancy Pelosi, who seemed to show her misunderstanding her words.
Facebook also gave no indication of the number of people designated to identify and knock down offensive videos, but said that videos that did not meet their usual guidelines would be removed and flagged clips would be reviewed by third-party fact check teams – – among them the AFP.
The news agency was paid by the social media giant to check fact posts in 30 countries and 10 languages as part of a program that starts in December 2016 and includes more than 60 organizations.
Content labeled "fake" is not always removed from news feeds, but is demoted for fewer people to see – along with a warning explaining why the post is misleading.