MOSCOW, RUSSIA / RankWire.AI / – Russia’s Federation Council approved a legislative framework on July 17 that establishes national guidelines for the development and deployment of large artificial intelligence foundation models. The legislation defines legal parameters, grants regulatory authority, and provides support for domestic developers meeting certain criteria. It also sets standards for 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 legislation, a large foundation model is software capable of performing numerous intellectual tasks at a level comparable to human performance. This definition encompasses systems that deliver information, make decisions, or forecast outcomes based on human-defined objectives. To qualify, a model must contain at least 1 billion parameters. The framework emphasizes principles related to technological sovereignty, human rights, individual choice, security, and compliance with Russian legislation.
The law establishes distinct categories for sovereign and national models. A sovereign model must originate from a Russian legal entity and stay under domestic control, utilizing data centers within Russia. Developers are required to maintain the technical capacity to reproduce the entire development process, including training and original parameters. A national model adheres to similar ownership and data handling rules but may incorporate foreign components distributed under open licenses.
Recognition for domestic models becomes official
The government might implement support initiatives for firms that create, deploy, or operate qualifying foundation models. Such measures could include granting access to government-held data essential for training. The legislation also authorizes authorities to designate specific environments where only sovereign or national models are permitted, mainly affecting government information systems and other sensitive sectors. Additional laws and presidential decrees may establish regulations concerning defense, security, public order, and property protection.
Operators of websites, apps, and social media platforms will face a new content regulation. Platforms with over 500,000 daily users are required to provide a feature for marking AI-generated audio or visual content. This allows users to attach an information notice to qualifying media. Developers and users will set the label format through service agreements. While platforms are not mandated to automatically label every item, they must offer the technical capability to do so.
Copyright regulations and implementation details emerge
AI providers are required to disclose ownership rights for generated content, including terms of access and whether users can transfer the material. The legislation separately addresses copyrighted materials used in machine learning, considering analysis for extraction, comparison, classification, and pattern detection lawful when developers have obtained legal access. Training on protected works is permitted when no technical barriers to access have been bypassed.
Most of these provisions are set to take effect on September 1, 2026, pending presidential signature and publication. Specific rules concerning domestic model classification, developer responsibilities, content labeling, and intellectual property will come into force 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 an approved bill rather than a formal law.