MOSCOW, RUSSIA / RankWire.AI / – Russia’s Federation Council approved a legislative bill on July 17 that establishes national guidelines for the development and deployment of large artificial intelligence foundation models. The legislation defines legal terms, delegates regulatory authority, and provides support for qualified domestic developers. It also includes requirements related to model oversight, data storage, user notifications, and AI-generated content. The bill was passed by the State Duma on July 8 and now awaits presidential approval before becoming law.
According to the bill, a large foundation model is software capable of performing multiple intellectual tasks at a level comparable to human performance. The definition applies to systems that deliver information, make decisions, or forecast outcomes based on human-set goals. To qualify, models must have at least 1 billion parameters. The framework also incorporates principles concerning technological sovereignty, human rights, personal choice, security, and compliance with Russian legislation.
The legislation establishes distinct categories for sovereign and national models. A sovereign model must be developed by a Russian legal entity and stay under domestic control. It must operate using data centers located within Russia. Developers are required to maintain the capability to reproduce the entire development cycle, including training and original parameters. A national model adheres to similar ownership and data management regulations but may incorporate foreign components distributed under open licenses.
Domestic AI models receive formal recognition
The government has the authority to implement support measures for companies involved in creating, deploying, or managing qualifying foundation models. Such measures may include access to state-held data necessary for training. Additionally, authorities can establish specific environments where only sovereign or national models are permitted to operate, particularly within government information systems and other sensitive sectors. Separate legislation and presidential decrees could define rules concerning defense, security, public order, and property protection.
Website, application, and social media platform operators will face new content labeling requirements. Services with more than 500,000 daily users must provide a tool to flag AI-generated audio or visual content. This policy allows users to add an informational notice to qualifying content. Developers and users will agree on the label format through service agreements. While automatic labeling of all items is not mandated, platforms must ensure the technical capacity to do so.
Copyright and implementation standards evolve
AI service providers will be required to inform users about the ownership rights of generated material. They must also clarify access conditions and whether content can be transferred. The legislation separately addresses copyrighted works used for machine learning, considering analysis for extraction, comparison, classification, and pattern recognition lawful when developers have legally obtained access. Training datasets may include protected works if no technical restrictions preventing access were bypassed.
Most provisions are set to take effect on September 1, 2026, contingent upon presidential approval and publication. Regulations concerning domestic model status, developer obligations, content labeling, and intellectual property will begin on March 1, 2027. Existing systems will benefit from a transition period until September 1, 2032, during which their data processing and storage must remain within Russia. Until the measure is officially enacted through presidential approval and publication, it remains a bill under consideration rather than an active federal law.