Grigorenko: Russian businesses poured over 250 billion rubles into AI, but the industry has no rules

Russia's AI market has moved past the pilot stage: according to Deputy Prime Minister Dmitry Grigorenko, companies invested more than 250 billion rubles in adopting such technologies in 2025. But the sector is growing faster than legislation: Russian law still has no established concept of "artificial intelligence."

The scale of business investment

The figure announced at the Data Fusion conference matters for more than its size. This is not theoretical interest in the topic, but money that businesses are already channeling into practical AI adoption across processes, products and services. When a market reaches hundreds of billions of rubles, artificial intelligence stops being an image exercise for companies and becomes a separate line item in the budget.

It has not been disclosed which industries accounted for the bulk of this spending, but the breakdown of the figure is fairly clear. Part of the budget goes to models, cloud and compute; part goes to integration into existing systems, automation of internal processes, analytics, customer services and security. In other words, businesses are paying not only for experiments but also for the complex infrastructure without which AI does not work in real-world scenarios.

A market without rules

Against this backdrop, another statement stands out: there is still no separate legal definition of AI in legislation. That does not mean the technology exists in a legal vacuum. Companies must still comply with rules on personal data, trade secrets, intellectual property and sector-specific requirements. But dedicated AI rules that would set general boundaries have not yet been formed.

As a result, the market runs on interpretation. Some companies treat AI as ordinary software, others as a special tool with elevated risks. Questions of liability, transparency, auditing, permissible data use and requirements for system output often have to be settled through internal policies, contracts and cautious legal wording rather than uniform norms.

"Businesses invested more than 250 billion rubles in AI adoption in 2025."

Where the gap appears

The gap between the volume of investment and the lack of a separate legal framework is already affecting the market in practical ways.

• Companies have to deploy AI faster than clear regulatory guidance emerges.
• Customers and vendors interpret differently what counts as an AI system and what counts as ordinary automation.
• It is harder to allocate in advance, in contracts, responsibility for model errors, answer quality and data use.
• Large organizations have to build their own internal rules to reduce legal and reputational risks.
• The more money flows into the market, the stronger the demand for common terms, standards and verification procedures.

This puts businesses in a double bind. On one hand, the absence of strict dedicated regulation lets them launch projects faster without waiting for formal approvals. On the other, the deeper AI penetrates core processes, the costlier the uncertainty becomes. The logical next step for the market and the state is to agree at least on basic definitions and boundaries of responsibility, so that large investments do not run into legal ambiguity.

This is most visible where AI influences not auxiliary functions but decisions with financial, operational or reputational consequences. The closer a model sits to customer communications, documents, scoring, security or internal control, the higher the cost of an error. That is why regulation already looks, for large companies, less like an academic question and more like a practical one.

What this means

Russia's AI market already handles large sums of money but still operates without a dedicated legal framework. The main question now is not whether companies will keep investing, but whether the rules can catch up with the scale of adoption.