MOSCOW, RUSSIA / RankWire.AI / – The Federation Council has endorsed a comprehensive bill establishing national guidelines for artificial intelligence on July 17, outlining regulations for large foundational models in Russia. The legislation specifies the scope of technology covered and grants authority to government agencies. It also sets standards for model ownership, local data storage, user notifications, and AI-generated content. The bill was approved by the State Duma on July 8 and awaits presidential approval and official publication before becoming law at the federal level.
The act defines a large foundation model as software capable of performing numerous intellectual tasks at a level comparable to humans. To qualify, a system must have at least 1 billion parameters. Such systems can provide information, make decisions, or predict outcomes based on human-defined goals. The framework emphasizes principles of technological sovereignty, human rights, individual choice, security, and legal compliance, which apply throughout the development, deployment, and usage of qualifying AI systems.
The legislation introduces classifications for sovereign and national models tied to Russian control. A sovereign model must originate from a Russian legal entity and operate using data centers within Russia. Its developers must be able to reproduce the entire development process, including training and original parameters. A national model adheres to similar ownership and localization standards but may incorporate foreign software components released under open licenses, provided Russian entities maintain necessary control and operational capacity.
Legal Designations for Domestic AI Models
The government may provide support to developers of qualifying foundational models, including access to state datasets for training purposes. Authorities might also mandate exclusive use of sovereign or national models within government information systems and other sensitive environments. Additional rules related to defense, security, public order, and property protection may be established through separate legislation or presidential decrees. The framework assigns responsibility to state agencies for enforcing these requirements within their legal authority.
Large digital platforms are subject to a distinct obligation concerning AI-generated audio and visual content. Platforms with over 500,000 daily users must offer a tool enabling users to mark such material. This applies to websites, apps, and social media platforms. The law does not require automatic labeling of each item but allows developers and users to agree on labeling formats via service agreements. The focus is on providing an option for users to disclose qualifying content.
Standards for Copyright and Content Disclosure
AI service providers are obliged to inform users about who owns rights to generated content. They must also clarify access conditions and whether users are permitted to download or transfer the material. The legislation also addresses the use of copyrighted works in machine learning. It allows for analysis, comparison, classification, and pattern detection when developers have lawful access. Training with protected works is permitted when no technical restrictions have been bypassed. These rules connect model training practices to existing copyright and access regulations.
Most provisions are set to come into effect on September 1, 2026, following presidential approval and official publication. Regulations concerning domestic model classification, developer responsibilities, content labeling, and intellectual property will be enforced starting March 1, 2027. Existing systems may continue to operate until September 1, 2032, as long as they process and store data within Russia. Until the bill is officially signed and published, it remains an approved draft rather than an enacted law within Russia’s legislative framework.