Facebook is still trying to limit borders truths and lies ahead of the elections this year in the united States, and puts the gaze on the so-called deepfakes, videos, fake hyper that use techniques of artificial intelligence to edit the physical features of a person and confuse it with another.
In a policy update announced on the Monday night, the vice-president of management of global policies of the social network, Monika Bickert, said he will take a position that is more strict on the content manipulated in the media from here on out, removing content that has been edited or synthesized "in a way that are not obvious to the average person." Bickert, however, noted that the issues of quality or cuts and splices in videos that simply reduce or change the order of the words are not covered by the ban, nor the content parodic or satirical. The material will be removed if it is "a product of artificial intelligence or machine learning which overlays or replaces content in a video, making it look authentic," says Bickert.
MORE INFORMATIONThe ‘deepfakes’ complicate the fight against the false news reports
it will Also remove the videos created from AI or machine learning that merge, replace, or overlap another video to make it look authentic.
The company was already working on the identification of deepfakes since September of 2019 with his Deep Fake Detection Challenge which has united people from all over the world in the search of the contents manipulated. In addition, we have joined Reuters to create a training course to help journalists to identify false news.Updated Date: 07 January 2020, 19:00