The Russian State Duma has passed in first reading a law on the… (модель · законопроект · технология)
The Russian State Duma has passed in first reading a law on the development of artificial intelligence technologies, establishing classifications for sovereign and national AI models, regulating content creation with AI, and providing mechanisms for state support. The legislation aims to balance technological advancement with national security and digital sovereignty.
Consensus
- The Russian State Duma has passed the law on the development of artificial intelligence technologies in first reading.
- The law introduces two categories of large fundamental AI models: sovereign and national.
- Sovereign AI models must be developed exclusively by a Russian legal entity, fully controlled within Russia, and hosted in data centers located on Russian territory.
- National AI models must also be developed by a Russian legal entity but may use foreign components under open licenses.
- Both model types must comply with Russian legislation and 'traditional spiritual and moral values'.
- The law regulates the development, deployment, and application of AI programs, including both domestic (YandexGPT, GigaChat) and foreign (ChatGPT, Gemini, DeepSeek) systems operating within Russia.
- The law addresses copyright issues by exempting from infringement claims the use of copyrighted material for training AI models if obtained legally or in open access.
- AI-generated audio-visual content must be marked with a warning, though the obligation is currently non-mandatory and will be implemented gradually.
- The government has the authority to provide financial, property, guarantee, and informational support to developers.
Points of divergence
- The requirement for AI content marking becomes mandatory on March 1, 2027, while the law itself enters into force on September 1, 2026. — interfax
- The format and order of AI content marking will be determined through agreements between the user applying the model and the platform provider. — fontanka
- The Ministry of Digital Development previously stated that AI in Russia should be censored, and a draft law was published earlier to regulate AI use with emphasis on protecting citizens from manipulation and discrimination. — kommerant
- The government plans to implement 40 AI-based services, aiming to accelerate data processing and improve procedural decision-making efficiency. — vedomosti
- The law allows for a balanced approach by distinguishing between sovereign (fully domestic) and national (incorporating foreign experience) models, ensuring both access to global advancements and data security. — vesti
Coverage (12 sources)
- Deputy: Russia's AI law allows maintaining balance in technology development — Вести
- Senator Pushkov: AI laws are needed to protect the population from dehumanization — Вести
- Grigorenko: Proper implementation of AI processes requires correct perception — Вести
- Grigorenko Explains the Need for Legal Regulation in the AI Sector — Вести
- AI Law Will Protect Russians from Risks — Вести
- The State Duma expanded the circle of internet platforms required to label AI content — Интерфакс
- State Duma adopts law on the legal basis for developing artificial intelligence — Вести
- "United Russia" Will Ensure Support for Developers of Domestic AI Models — Вести
- AI by statute: AI regulation law adopted in first reading — Фонтанка
- The State Duma approved the AI regulation project in first reading — Коммерсантъ
- The State Duma approved the AI development bill in first reading — Ведомости
- The State Duma will consider the AI development bill in first reading — Интерфакс