Skip to content
#AI

Artificial intelligence in Russia 2025: trends and prospects (AI column)

#AI #Engineering #Metrics #Software #Productivity #Economics

This weekend I read the full article carefully. report, which was prepared by the consulting company "Yakov and Partners" together with Yandex (Business perspective + technology expertise). The report relies on three data sources: survey 150 Technical Directors of Large Companies (from 16 industries)poll 150 vendors of AI solutions, survey 3500+ Russians. In-depth interviews with industry leaders were also conducted to look below the surface of the numbers. A brief review of the results is landing organizers, where there are quotations of the type

According to the SRT survey, over two years, generative AI has gone far beyond point experiments: the average number of functions where pilots are running or full implementation has grown from one to the next. 2,4 into 2023 before 3,1 into 2025The technology itself is already being used in 80% of key business functions

I have already mentioned these quotes in my post.Leadership surveys and relationship to reality"but today I'm going to talk about the overall structure of the report and what I found interesting:" The authors explored four key areas of AI development. 2025 genAI, NLP & Speech, CV, RecSys. These technologies are used in companies and are drivers of growth. The report outlines trends in each of these areas.

  • The structure of the report is remarkable. First, a picture of the general state of the AI market in Russia and the level of technology implementation is given, with an assessment of the economic effects of AI implementation. The following are detailed chapters on each 4 Technology: where to use, how to progress, which cases are most popular In conclusion, the authors share a common sammmari and their view of the future, as well as provide recommendations for everyone. The main conclusions can be drawn from the conclusion and I will try to tell them thesis.

Generative AI is a horizontal platform GenAI is no longer seen as a separate technology, but as a horizontal framework that permeates and transforms all other areas. (NLP, CV, RecSys). GenAI is becoming a universal environment on the basis of which you can solve a variety of tasks.

The era of foundation models and agent systems. The market is entering a new phase where basic models dominate the scene. (foundation models) And autonomous AI agents. Today’s big models are no longer passive repositories of knowledge – they are becoming active performers. There are protocols of interaction between them, and instead of the usual applications, we will increasingly work with smart services that act on our behalf.

Barriers to the implementation of AI already on the “iron” If earlier the main restrictions were technology and access to capacity, now the key barriers are organizational and infrastructure. It is not the missing algorithms that hinder the implementation of AI, but the cost of scaling. (infrastructure, inference, restructuring of processes) Conservatism of companies and users.

The future of AI in Russia according to the authors: the prospects are very optimistic The Russian AI ecosystem is entering a scaling phase. It is estimated that 2030 The effect of AI for the Russian economy can reach 7,9–12,8 trillion per year (≈5,5GDP) This is comparable to the profits of the entire banking industry. At the same time, it is not only about reducing costs, but also about revenue growth due to new products and personalization of services. The authors believe that in a few years, the introduction of AI will be a matter of survival for most companies.

In the end, the authors gave advice to everyone.

  • Business. Build unified AI platforms for models/data/tools, where possible in the cloud for flexibility. Focus on several priority implementation cases and soberly plan the effect (12–24 We need to pay off, not wait for miracles tomorrow.).
  • State - Expand support measures for AI initiatives: grants, R&D subsidies, tax breaks and affordable financing.
  • Users Keep up: systematically master AI tools in study and work, pump data skills and critically evaluate the results of models.

#Engineering #AI #Metrics #Software #Productivity #Economics