The social network has just modified its conditions of use in order to permanently ban fake videos.
In recent months, the popularity of deepfakes , these extremely realistic fake videos, has grown steadily. So much so that, even if they are sometimes used to make jokes that are not always very tasteful, many people take them for cash and share them, thus contributing to the spread of false information, especially on social networks.
The deepfakes can indeed be used to say anything to celebrities or political figures , or to change the head of adult film actresses .
After having recently launched its Deepfake Detection Challenge to try to advance research in the detection of deepfakes , Facebook is spending the second and has just modified its conditions of use.
Avoid disinformation before the presidential elections
These new rules, put in place to prevent the spread of false information before the US presidential elections, concern all videos that have been modified through the use of artificial intelligence or machine learning. However, not all videos should be affected by this filtering.
Facebook indicates that satirical and parody videos, or content modified to simply delete or change the order of words spoken by a protagonist, will not be affected. They may however be checked in order to limit their spread in the Internet news feed.
The social network also explains that its strategy is not to simply delete these videos: