Videoconferencing in Russia will be protected from deepfakes

Video conferencing will close loopholes for deepfakes. Domestic services — IVA Technologies, MTS Link and Vinteo — are completing the implementation of deepfake detectors, the companies told Izvestia about this. The technologies being created are relevant in the light of the massive spread of deepfakes, when fraudsters impersonate, for example, company executives, government officials and heads of regions.

False challenge: every second Russian will face a deepfake attack by the end of the year

How fake video and voice messages are created

These technologies are capable of detecting fake videos and audio with over 98% accuracy. New solutions are expected to make online communication more secure and reliable. Fraud will be detected automatically, reducing the risk of fraud. The technologies being developed use artificial intelligence and machine learning to analyze audio and video streams.

The algorithms pay attention to subtle signs of falsification: image distortion, out-of-sync lip movements and speech, features of synthetic voice, told MWS AI. As a result, the system can quickly and accurately determine whether the person you are talking to is real or fake.

"The product will be relevant for cloud video conferencing services. This is especially important given the increase in cases of fraud using social engineering methods and identity substitution attempts. The technology is currently undergoing a stage of refinement and testing," Maxim Smirnov, Deputy General Director of IVA Technologies, told Izvestia.

MTS Link, in turn, has already completed testing of this technology, said Dmitry Kryukov, head of the company's machine learning department. Roman Samoilov, CEO of Vinteo, noted that against the background of regular information attacks, the approach to the security of VKS services is becoming more complex, and the fight against deepfakes is one of the elements of increasing security, which will develop in response to emerging requests.