More than 2 million rubles has been allocated to Roskomnadzor to filter internet traffic in the country using machine learning, according to a report by Forbes.
Roskomnadzor plans to introduce internet traffic filtering in 2026, using 2,27 million rubles, reported Forbes, citing the Roskomnadzor digitalisation plan which was presented for consideration by the presidium of the government commission on digital development, the use of technologies to improve the quality of life and doing business decisions on December 26.
Technology is due to be integrated into technical means of countering threats (TMC) which uses deep packet inspection (DPI) technology, which has already helped over 1 million sites to be blocked to date, at an average of 5,500 addresses per day.
Experts believed that the AI would allow Roskomnadzor to more effectively find prohibited content “not by internet addresses, but by words, expressions, sentences, or other signs”, identify mirrors of blocked sites and VPN services, as well as analyse encrypted traffic.
Roskomnadzor are seeking to find prohibited content using machine learning technologies and identify VPN services, confirmed a Forbes source close to the office of Deputy Prime Minister Dmitry Grigorenko, although Roskomnadzor didn’t provide any official comments.
According to Forbes, Roskomnadzor already uses AI and automated systems Oculus and Vepr for monitoring text, audio and video, reducing the time to detect violations from 48 hours in 2020 to 6 hours in 2025.
However, not all systems have shown high efficiency of neural networks, because in some cases, the accuracy was around 60%, wrote Forbes.








