Russia is developing a system for automatic detection of prohibited information, “Oculus” will analyze photos, video images, texts on websites, social networks and messaging services, according to Moscow mass media cited by Rador.

Roskomnadzor, Russia’s media and communications watchdog, has already allocated 57.7 million rubles to develop the system, according to Kommersant, Russia’s largest daily financial newspaper.

A private company will be engaged in the creation of “Oculus”, the project is planned to be completed by December 12, 2022.

The system must detect prohibited content in real time not only in the form of text, but also in “sequences, combinations of objects, image compositions, faces, as well as static and dynamic movements.”

The capacity of the system should be 200,000 images in 24 hours, that is, Oculus will analyze two frames or images every second.

Kostyantyn Bulanov, Director of Digital Technologies of Roskomnadzor, explained that at least 48 servers equipped with graphics accelerators will be needed for the stable operation of the system.

Russia considers any information not approved by the Kremlin to be “fake”.

Bulanov told the RBC channel that Oculus does not independently search and collect information on the Internet. “Other information systems have been developed for monitoring, which are already working as part of the single analysis module (EMA),” he noted.

Data from monitoring systems will be uploaded to Oculus automatically.

Since the first half of March, the Russian parliament has adopted a draft law that provides for criminal liability for the dissemination of information about the country’s armed forces, which the Moscow authorities classify as “false.”

The scope of the law was later expanded to include information on all of its agencies operating abroad.

Numerous activists, dissidents or journalists have already been prosecuted or fined under the law, which also provides for administrative fines in cases the authorities deem less serious.