MOSCOW, March 18 – A new legislative draft published for public review outlines a sweeping data localization requirement for artificial intelligence used in Russian state infrastructure. The proposal mandates that only AI models processing data exclusively within Russia's borders can be deployed in government information systems and critical infrastructure facilities.

The bill establishes a 'trusted AI models' registry. To be listed, an AI system must meet three core criteria. First, it must guarantee all data processing occurs on Russian territory. Second, the model must satisfy security standards set by federal authorities responsible for countering technical espionage and information protection. Third, it must meet quality requirements defined by relevant federal agencies, state corporations, and the Central Bank.

The government, acting on proposals from sector-specific federal bodies, will define which critical infrastructure objects must use these registered, domestic-only AI systems. The Ministry of Digital Development indicated the law is slated to take effect on September 1, 2027. Officials argue the framework will create clear rules for developers, businesses, and the state while protecting citizens from hidden algorithmic manipulation and discrimination. The move signals a decisive step toward technological self-reliance in sensitive sectors.

Source: RIA Novosti