71% of large Russian companies already use generative AI in at least one business function. This is stated in the study "Artificial intelligence in Russia — 2025: Trends and Prospects", prepared by the consulting company "Yakov and Partners" and Yandex.
The study describes the state of the Russian artificial intelligence market, evaluates the economic impact of the introduction of AI and highlights key trends. The research is based on the results of surveys and in—depth interviews. 150 technical directors of large Russian companies from 16 industries, 150 suppliers of AI solutions and over 3,500 Russian users took part in them.
AI Penetration
The introduction of artificial intelligence is uneven. In some industries, it has already become a key factor for growth and efficiency improvement, while in others it still remains a tool for targeted improvement of individual business processes.
- Companies from five industries are most actively using artificial intelligence: IT and technology, telecom and media, e-commerce, banking and insurance. These industries have already moved from pilot projects to the systematic application of AI in business tasks.
- Over the past year, companies from advanced industries have allocated from 13 to 17% of their annual IT budget to the development of AI-based solutions. In the market as a whole, this indicator is lower — about 11%. Businesses are investing the most in generative AI, and next year they plan to increase this expense item by 25%.
- The vast majority of the surveyed technical directors and vendors — about 90% — believe that the level of development and implementation of artificial intelligence in Russia is above the global average or comparable to the level of the leading countries. At the same time, the level of development of AI solutions is rated higher than the level of their implementation.
- By 2030, more than half of the world's population will use generative AI.
Economic effect
In industries where AI is systematically used, the economic effect of its implementation is estimated at up to 8% of the company's EBITDA. AI allows you to reduce costs and gain new sources of revenue through product innovation and business model transformation.
- 78% of companies noted that they receive an economic effect from the introduction of artificial intelligence, which is 10 percentage points more than in 2023. At the same time, almost every tenth company records it at the level of 5% of EBITDA — two years ago such cases were rare.
- Companies from advanced industries estimate the cumulative economic impact of artificial intelligence at up to 8% of EBITDA and predict that in a year it will be from 13 to 21% of EBITDA. In other industries, expectations are limited to the range of 5-10% EBITDA.
- 90% of respondents say that the financial effect of AI is primarily due to a reduction in operating and other costs. 25% also note revenue growth due to the creation of new services and products. In advanced industries, this share is much higher — 70%.
- The expected economic effect of AI by 2030 is estimated at 7.9–12.8 trillion rubles per year, which is up to 5.5% of the projected GDP. It will be driven not only by cost optimization, but also by revenue from product innovation and business model transformation. The largest contribution of all AI technologies will be made by generative AI - from 1.6 to 2.7 trillion rubles.
- The maximum specific effect on EBITDA is in industries where AI transforms the business model and creates new sources of monetization. These are e-commerce, telecom and media, IT, medicine and healthcare, construction and real estate.
Technologies
Most companies use ready-made solutions from external contractors using cloud and hybrid models. In the field of generative AI, open-source models are most often chosen and adjusted to suit themselves without significant architectural changes.
- 78% of companies use ready-made AI solutions from external partners. About half of the respondents note that they adapt these solutions to their business processes together with the contractor. Only 17% of companies develop their own AI solutions.
- The most common delivery model is cloud—based: 40% of respondents use it. 29% choose a hybrid model. The situation depends on the industry: for example, in the banking sector with high data security requirements, 90% of companies use on-premises.
- Companies working with generative AI use mostly open-source models. This accelerates the introduction of solutions to the market, reduces barriers and promotes the mass adoption of AI. Instead of a full training cycle, 71% of companies choose to customize ready-made models for their tasks without drastic changes.
Trends
Generative AI complements and enhances other technologies, such as recommendations. The list of scenarios for its application is constantly expanding due to the development of multimodal models and AI agents, as well as the introduction of robots and smart devices.
- There is a transition from monomodels to general-purpose multimodal models capable of processing heterogeneous data. They cover a wide range of scenarios and identify key structural shifts in the coming years.
- Generative AI enhances other AI solutions, including recommendation systems, computer vision, and speech technologies. For example, it allows you not only to improve the quality of recommendations, but also to make communication with users more personalized.
- Artificial intelligence is gradually moving into the physical world, integrating into robots and wearable devices such as headphones or voice recorders. This allows systems to perceive their environment, make decisions, and act accordingly.
- A class of AI agents is actively developing, which expand the scenarios for using artificial intelligence. This allows you to move from questions and answers to multi-step autonomous activities and integration with corporate systems.
