The deepfake consists of a new technique based on artificial intelligence that allows you to superimpose the face of one person on another in a video, distorting reality, and making something that does not seem real, such as a speech. of a politician who perhaps said something else entirely.
Meta has announced the removal of all deepfakes from its platform precisely because these videos are very dangerous and are also used by criminals by showing videos that talk about untrue things, even if it seems that it is really difficult today to be able to distinguish a real video from a fake one.

The deepfake technique allows, through the exploitation of artificial intelligence, to superimpose the face of a person on another shot in a video. Then, through deep learning, the real faces of the two people are replaced (face swapping), thus creating videos in which you can make people say what you want, even with the same real voice and synchronized lip movement. This is how we find ourselves faced with a fake or fake video that we cannot distinguish from the real one.

All this leads to an enormous manipulation of information, also generating situations of panic among users as happened during the Covid pandemic where videos were put online that stated completely false things, creating false alarmism.
The most famous case of deepfake is certainly the video of Barak Obama who went around the world or the fake interview with Putin, but, to tell the truth, many famous people are victims of this very dangerous video. Defending against deepfakes is very difficult for the simple fact that fakers are very avant-garde and are very skilled at copying videos and editing them to their liking.