Russian authorities have explored the use of artificial intelligence to identify and block VPN services. The initiative, reportedly proposed by the Ministry of Digital Development, aimed to achieve a 95% efficiency rate in blocking such services by 2030.
The proposed plan involved using AI to analyze 70% of all blocked VPN services. The technology would automate the analysis of connection patterns, such as packet size, rhythm, session duration, and traffic ratios, to detect VPN servers based on indirect indicators.
The Ministry of Digital Development stated that the current version of the national AI implementation plan does not contain these specific provisions. Experts note that authorities already employ traffic analysis algorithms, though these are not currently classified as AI.
Industry experts warn that relying on AI for blocking carries risks, including the potential for high false-positive rates. There is concern that legitimate corporate traffic, cloud access, and voice data could be incorrectly identified as VPN traffic, causing significant business disruptions.
To mitigate risks, experts suggest using AI only as a supplementary tool rather than for autonomous decision-making. A proposed workflow involves AI identifying suspicious nodes, followed by manual verification by specialists before any blocking actions are taken.
In addition to AI initiatives, other measures are being considered to restrict VPN usage. A proposed regulation would prohibit hosting providers from supplying capacity to clients running VPN services for one year, with offenders being added to a Roskomnadzor registry.
