The Moscow authorities will spend 130 million rubles to finalize the facial recognition system used by the police. It is assumed that they will be able to use city video cameras and neural networks not only to find criminals, but also to calculate their movement routes and alleged accomplices. The system will also recognize people who often come to the scenes of crimes committed. According to experts, not only suspects and their accomplices can get into the field of view and the neural network database, but also ordinary citizens whose data can later leak into the network.Kommersant found on the public procurement portal that the Department of Information Technologies (DIT) of Moscow on August 27 announced a tender worth 130 million rubles for the modernization of the subsystem for automatic registration of video information indexing scenarios (PARSIV). This is an IT solution integrated with the Unified Data Storage Center (ECCD) of the Moscow City Hall. PARSIV, in fact, is a gateway through which police officers can connect to the city’s facial recognition system to search for criminals. The contractor will be chosen on October 1, 55 days are given for the work to be completed.According to the technical documentation, the police, for example, will be able to search not only for suspects, but also for their accomplices: the neural network will track the movement of the criminal, and then record who often appeared next to him and at what time. It will also be possible to track people who return to those crime scenes “where the identity of the criminals has not been established.”Advanced filters will be added to the PARSIV to search for people: by gender, age, race, the presence of glasses, beard, mask, headdress and the ability to recognize silhouettes. In addition, the system will allow the police to inspect the route of the object at a given site.Finally, the tender involves the development of a mobile application for the iOS and Android operating systems, with which police officers will be able to connect to the face recognition system from tablets and smartphones. “New tools will appear in PARSIV, which will allow to ensure the safety of citizens even more effectively, and the security of the subsystem itself will also be increased,” the press service of the DIT noted. According to the agency, today in Moscow, video analytics algorithms process video streams from 125 thousand cameras. The Interior Ministry did not respond to the request.In total, the Ministry of Internal Affairs plans to spend 55 billion rubles on its own digitalization and modernization of IT systems (see Kommersant of November 16, 2020). The Ministry has already ordered a study of machine learning technologies that will help solve serial crimes by identifying the handwriting of one criminal in different criminal cases (see Kommersant of August 30).Technology for tracking people, including video-based inter-camera tracking.�лит lithiki, appeared in Russia a few years ago, it was successfully used during the 2018 FIFA World Cup, said Alexander Rozhkov, Director of Products and Technologies of T1 Group. According to him, the method was widely used last year during the period of self-isolation.Artem Romanov, an expert on security systems at Krok, notes that now the face recognition system in Moscow is used to determine the static position of a person: it allows you to find out where a certain person was at a certain point in time. The next step, in his opinion, is an automatic multidimensional analysis of data: “Where else did you see a particular person today or who else was seen in the right place during the week.” Now, the expert notes, only an operator can do this and it will take a lot of time: “The goal of the system being created is to get the result in minutes by setting the necessary parameters.”But not only suspects and their accomplices can get into the field of view and the neural network database, but also ordinary citizens, warns the expert of Roskomsvoboda, senior lawyer of the Digital Rights Center Vladimir Kolerev: “The presence of a centralized database with a large amount of data about citizens makes it a convenient target for attackers.” Even with the use of information security tools, experts fear, the data can get into the black market.Nikita Korolev