According to a 2021 report from the U.S. Department of Homeland Security, the threat of deepfakes does not come much from the technology used to create them, but from people’s natural inclination to believe what they see.
Certainly, the big players in the technology sector such as Google, Microsoft and Meta have openly condemned deepfake technology and are concentrating their efforts to create increasingly performing tools to identify these altered videos. Microsoft, for example, is creating a new anti-deepfake technology to fight disinformation better known as the Microsoft Video Authenticator.
Despite our efforts, however, we are nowhere near achieving an effective system capable of distinguishing real audio / video content from fake ones. Some of the best tools available are Counter.social, Deeptrace, Reality Defender, and Sensity.ai. The latter, according to some rumors, seems to be the first authentic deepfake detection in the world.
The effectiveness of the differentiation system between true and false content represents the element on which the efforts of all companies committed to the purpose are focusing. This is essentially due to the current level of reliability of the platforms currently available, which stands at just over 65%. Even Azure Cognitive Services were fooled 78% of the time.
Even Rijul Gupta, co-founder of DeepMedia, based in Oakland, California, in a recent statement released regarding Deepfake’s unstoppable expansion, admitted: “Everything will be able to be faked in real time”. “