MMOOZZGG.com

AI for lawyers

AI regulation in Russia: what a lawyer should know in 2026

5 min read10 sources

AI regulation in Russia in 2026 has five layers: law 243-FZ, the National Strategy, the AI Code of Ethics, the personal data law 152-FZ and court decisions. Federal Law No. 243-FZ of 26 July 2026 came into force on 1 September 2026, but its main rules on models start only on 1 March 2027. Below is a map of these documents for a lawyer who advises clients and works with AI.

On this page · 5 sections

243-FZ: sovereign and national models

Law 243-FZ “On Support for the Development of Artificial Intelligence Technologies in the Russian Federation” is a framework law. It regulates large foundation models: under Article 3, this is a program that serves as the basis for other software, has at least 1 billion parameters and solves many different tasks.

The central idea of the law is sovereign and national models. The data of such models is processed in Russia, and the government defines the cases where only these models may be used.

Article 13 sets different dates. The law has applied since 1 September 2026, while Clauses 3–5 of Part 2 of Article 5, Parts 2–5 of Article 6 and Articles 8–10 (developers’ duties, labeling of content, rights to results) come into force on 1 March 2027.

For systems already running on the day Clause 3 of Part 2 of Article 5 comes into force, the limits on the choice of model do not apply until 1 September 2032, if the data is processed and stored in Russia. Hence the first question to a client: where is the data in its AI systems processed today?

What to check at a client before 1 March 2027:

1. Which models run in its systems and where they are deployed.

2. Where the data that passes through the model is processed and stored.

3. Whether the model can be replaced without rebuilding the system if the government limits the choice for its sector.

What the law changes for companies that simply use AI services is covered in the article “AI law 243-FZ: what business should do”.

National Strategy

The industry’s strategic document is Presidential Decree No. 490 of 10 October 2019 “On the Development of Artificial Intelligence in the Russian Federation” and the National Strategy for AI Development to 2030 that it approved. The current version was set by Decree No. 124 of 15 February 2024.

The strategy is a strategic planning document. It sets no direct duties for companies, but later acts and industry documents rely on it.

An example of such reliance: in Section 2 (Clause 1.1), the Code of Ethics names the National Strategy among its legal foundations, together with the Constitution, the National Security Strategy and the Concept for Regulating AI and Robotics.

The same direction is fixed in Article 4 of law 243-FZ: technological independence, support for Russian developers, protection of human rights and legitimate interests, and a risk-based approach. In a dispute over how to interpret framework rules, these principles work as an argument about the purpose of regulation.

Code of Ethics

On 26 October 2021 Russia adopted the AI Code of Ethics. The document is a set of recommendations and covers the ethical questions of creating and using AI that laws and technical regulations do not settle.

The code’s recommendations apply only to civilian AI systems. Two provisions matter for a lawyer: a person, an individual or a legal entity, is responsible for all consequences of the system’s work (Clause 3.2), and human oversight and logs of the system’s work are part of Clause 3.1.

Joining the code is voluntary (Clause 2.1 of Section 2). An organization that joins appoints an AI ethics officer, and the register of participants is public: before negotiating with an AI vendor, you can check its status.

The code has no sanctions, but its wording is easy to carry into an AI use policy and into a contract with a vendor: the parties speak a language the market already knows.

Personal data

The main law in force for AI projects today is 152-FZ. The collection and storage of Russian citizens’ data take place in databases in Russia (Part 5, Article 18), and since 1 March 2023 an operator must notify Roskomnadzor, the data protection regulator, before a cross-border transfer (Part 3, Article 12).

For breaking the rule on databases in Russia, a legal entity pays a fine of up to $72,000 under Part 8, Article 13.11 of the Code of Administrative Offenses. A contract with full names and passport data pasted into a foreign chatbot raises both questions at once: where is the data stored, and was there a cross-border transfer?

The notice procedure and the risk assessment are covered on the page about cross-border data transfer, and the general link between the law and neural networks in the article “AI and personal data”.

In our projects, personal data is processed on servers in Russia under 152-FZ. The text does not go abroad or to services without a data processing contract.

Courts

Order No. 59-P of the Supreme Court of the Russian Federation of 3 September 2026 approved the Concept for introducing AI technologies into court proceedings. The areas: recording hearings, assigning cases, preparing draft decisions.

The principle of the concept: AI stays a tool of the judge, and a person makes the decision. The logic is the same as in Clause 3.2 of the Code of Ethics: the technology prepares the material, and a person bears the responsibility.

For a practicing lawyer, two things follow. AI will enter the work of courts, and whoever signs procedural documents is still responsible for their content; how courts react to made-up citations is covered on the page about court fines for AI.

A detailed analysis of the document is on the page “Concept for AI in court proceedings”. How Europe regulates AI is covered on the page about the EU AI regulation.

Sources and fact-checking

Checked by the editors on 16 October 2026.

  1. ppt.ru, 26.07.2026 (ppt.ru)
  2. ConsultantPlus, 243-FZ, Article 3, 26 July 2026 (consultant.ru)
  3. ConsultantPlus, 243-FZ, Article 13, 26 July 2026 (consultant.ru)
  4. Garant, 10 October 2019 (base.garant.ru)
  5. ConsultantPlus, 243-FZ, Article 4, 26 July 2026 (consultant.ru)
  6. AI Code of Ethics, official text, 26 October 2021 (ethics.a-ai.ru)
  7. AdIndex, 26.10.2021 (adindex.ru)
  8. Brace Law Firm (brace-lf.com)
  9. Data security in AI projects
  10. ppt.ru; Garant, September 2026 (ppt.ru)

Found a mistake? Tell us

Publications

Mike Mozg

Agentic AI use cases at AT&T: nearly 500 processes now run on agents

AT&T no longer uses AI only to help staff work faster: it is rebuilding whole business processes around agents. SiliconANGLE reported on 14 November 2026 that nearly 500 agentic AI use cases at AT&T are already in production. Here is how the company rebuilds them, where a human still checks the agent, and what other large companies can take from it.

News4 min
Contact us