Whoever hangs out on the web knows very well that today as the originality of a content is a huge subject; fake photos, false news, fake videos, the web is full of contents that have been manipulated, in a more or less professional way. A new trend in the world of online forgery concerns the manipulation of videos by means of a technique commonly called deepfake that allows the use of artificial intelligence to manipulate videos at a high level. This technique basically consists in gluing the face of a subject onto the body of another subject again, a bit like in the photomontage; in these cases, however, it is quite easy (at least at the current stage of technological development) to recognize this type of manipulation, which is very different when instead the deepfake technique is used only to alter a person's lip. Let's imagine for example the speech of a politician, through the deepfake it is possible to make the politician's lips utter a whole other discourse and, with the help of a professional dubbing, one can thus realize fakes very difficult to recognize; the effect, as anyone can understand, is potentially very dangerous since, for example, someone could put in circulation the videos of the CEO of a large company that declares bankruptcy, thus triggering panic on the markets without even the news being true.

Now, a recent report published by Witness Media Lab investigates precisely this type of threat and comes to identify the blockchain technology as one of the tools capable of defusing the mine represented by deepfakes videos; to tell the truth, for some time now many have argued that this is one of the most useful use cases of blockchain technology, but the Media Lab report is one of the very first cases in which this idea is written down and reasoned. The idea, however, is simple, does not deserve special explanations, it is simply a matter of storing content such as videos, photographs and audio files on a platform that "marks" them with a whole range of information ranging from timestamps to geolocation, all obviously protected from cryptography; however, even if the blockchain technology lends itself well to this type of use, there remain objective limits to all this, such as, for example, trivially, the fact that such a platform, to have any chance of working, should first establish itself as a standard.

In fact, it would be difficult to imagine such an infrastructure working without a sufficiently large amount of data; and even if everyone uploads their own videos and photos on the same platform nothing could prevent someone from making a video with the intention of using it later to produce a fake. In practice, therefore, through blockchain technology we might be able to define which of two videos is original, but only if the platform already has this information; it is not in fact a case that one of the greatest forensic computer experts, Hany Farid, has recently been able to state that any blockchain solution destined to counter the threat represented by deepfakes is still years away due to the complexity that a platform designed for this purpose.