MMOOZZGG.com

Glossary

What is Russia's AI law?

In short

Russia’s AI law is Federal Law No. 243-FZ on supporting the development of artificial intelligence technologies, which introduces the concepts of large foundation models, sovereign models and national models.

Security and law · AI law

MMOOZZGG.com editorsChecked: 1 min read

In depthRussia’s AI law 243-FZ for business: what to do before March 2027

For a child

At school there are rules: do not run on the stairs, listen to the teacher, do not take other people’s things. They were made not to get in the way but so that everyone is comfortable and safe. When smart computer helpers appeared, adults began to write rules for them too. Such a law says who may build smart programs and use them, and how.

In plain words

In Russia, the basic law on AI technologies was signed on 26 July 2026. It does not ban the use of neural networks. It describes which models the state supports and what their creators must do. It introduces the ideas of a sovereign model and a national model: a Russian company makes them, and answers and data are processed in Russian data centers. It also obliges large platforms to let people label content made with AI. Many details will come later in Government decrees.

For business

The law entered into force on 1 September 2026, and the rules on the status of models, the duties of their developers and labeling will start on 1 March 2027. Direct duties apply first of all to developers of sovereign and national models and to platforms with personal pages whose daily audience in Russia is more than 500,000 users. A company whose AI agent handles requests should check now where the data is processed, whether the model can be replaced and how the agent introduces itself to customers. A detailed review with dates and steps for business is in a separate blog article.

For a specialist

Federal Law No. 243-FZ of 26 July 2026 On Supporting the Development of Artificial Intelligence Technologies in the Russian Federation regulates the development, rollout and use of large foundation models. Under Article 3, such a model is a program trained on datasets, serving as a base for other software, used for many tasks and containing at least 1 billion parameters. Article 6 introduces sovereign and national models. Article 9 gives those who create audio and visual content with such models the option to post a notice that AI was used, and obliges large platforms to provide this option to users. Under Article 13, the law applies from 1 September 2026, and Articles 8–10 and the requirements of Article 6 from 1 March 2027. Limits on the choice of models do not apply to systems already in use until 1 September 2032, if the data is processed and stored in Russia.

Example

A chain of clinics books patients through an AI agent on a foreign service. To prepare for March 2027, it moves the processing of chats to a Russian data center, so that the coming limits on the choice of models do not catch it unprepared.

Related terms

Questions and answers

Does Russia have an AI law? Yes: Federal Law No. 243-FZ of 26 July 2026 On Supporting the Development of Artificial Intelligence Technologies in the Russian Federation.

It entered into force on 1 September 2026, and its main rules apply from 1 March 2027.

Do you have to label content made by a neural network? The law does not create a general duty to label all content. It gives authors of audio and visual content the option to post a notice that AI was used, and it obliges large platforms with personal pages to provide that option.

Sources

Not sure whether your company needs this?

AI solutions map
Contact us