WASHINGTON – Facebook said that it would ban videos that were heavily manipulated by artificial intelligence, the latest in a string of changes by the company to combat the flow of false information on its site.
A company executive said in a blog post posted late Monday that the social network would remove videos altered by artificial intelligence, often called deepfakes, in ways that “would likely mislead someone into thinking that a subject of the video said words that they did not actually say.” The videos will also be banned in ads.
The policy will have a limited effect on slowing the spread of false videos, since the vast majority are edited in more traditional ways: cutting out context or changing the order of words. The policy will not extend to those videos, or to parody or satire, said the executive, Monika Bickert.
Bickert said all videos posted would still be subject to Facebook’s system for fact-checking potentially deceptive content. Content that is found to be factually incorrect appears less prominently on the site’s news feed and is labeled false.
But the announcement by Facebook underscores how the social network, by far the world’s largest, is trying to thwart one of the latest tricks used by purveyors of disinformation before this year’s presidential election. False information spread furiously on the platform during the 2016 campaign, leading to widespread criticism of the company.