Neural network games: business is faced with a shortage of AI specialists
Russian business is faced with an acute shortage of artificial intelligence specialists. The demand for such employees significantly exceeds the supply: in the first 10 months of 2025 alone, hh.ru There were more than 1,000 vacancies for AI expert trainers who train corporate neural networks and evaluate their responses. Almost 45% of companies report difficulties in finding qualified personnel. The growing interest in AI specialists is directly related to the government's policy of strengthening technological sovereignty. Earlier, President Vladimir Putin stressed the need to form a national plan for the introduction of generative artificial intelligence.
How companies solve the problem with AI specialists
Russian business, despite the active introduction and financing of artificial intelligence technology, is facing an acute shortage of personnel. The demand for AI specialists is already many times higher than the supply, Izvestia found out. AI trainers, specialists who prepare training materials for AI models and evaluate the correctness, accuracy and ethics of their work, have become one of the fastest-growing professions.
Despite its relative novelty, the profession has rapidly gained ground in Russian companies. In the first 10 months of 2025 alone, employers have placed more than 1,000 vacancies for AI trainers, the director told Izvestia. hh.ru according to the research of Maria Ignatova.
The field of machine learning (ML) is developing no less dynamically. Since the beginning of 2025, ML specialists have been mentioned in more than 20,000 vacancies in Russia, and ML engineers have demonstrated the greatest increase in demand (74%), the expert emphasized.
The specialist market is not keeping up with the growth rate of the AI industry, and there is a particularly acute shortage of personnel with practical experience in implementing such solutions: those who know how to adapt technologies to business tasks and launch them into production, said Konstantin Mykhykh, Director of Data Science at Avito.
— The industry is young, many cases are being implemented for the first time. It is extremely difficult to find a specialist with relevant experience: few people have it. At the same time, the boundaries between roles are blurred — it's not enough to be an ML engineer, you need to understand product development, system architecture, and business specifics. There is a growing demand for hybrid competencies," he said.
According to him, Avito is now betting on growing its own staff.
Yandex is also actively developing the AI field: an average of 200 technical specialists go to the company for interviews every day, and ML engineers are traditionally at the top in terms of responses, the company's press service told Izvestia. They added that by the end of the year it is planned to hire ML engineers for 60% more than in 2024.
Boris Ashrafyan, director of the product and investment office and leader of AI-Native transformation at MTS Bank, confirmed the shortage of personnel in the market to Izvestia. According to him, today not only top managers with experience in major AI transformations are in demand, but also strong technical specialists at the middle and senior levels.
"The vast majority of companies have realized the need to develop expertise within the organization, but for this, the "core" is critically important — the carriers and conductors of knowledge and skills in the field of AI, these are people with rich experience," he believes.
MTS Web Services surveyed 700 companies with AI budgets: 44% lack specialists. The shortage of personnel is most acute in industry (60%), including heavy industry (69%), construction and housing (61%), science and education (59%).
Russian President Vladimir Putin speaks at the plenary discussion "The Future with AI" of the international conference on artificial intelligence "AI Journey" in Moscow, October 19.
The growing interest in AI specialists is directly related to the government's policy of strengthening technological sovereignty. President Vladimir Putin, speaking at the Sbera conference "Journey into the world of Artificial Intelligence" (AI Journey), noted that Russia cannot allow dependence on foreign neural networks. In this regard, the head of state instructed the government to create a headquarters for the management of all activities related to AI, as well as to form a national plan for its implementation.
How AI technologies are taught in universities
Today, education is gradually adapting to the speed of AI technology development, and cooperation with science is an important element of program flexibility, says Ivan Oseledets, Professor of the Russian Academy of Sciences and Director General of the AIRI Institute.
— I support a network model in which practitioners and scientists are involved in educational programs in order to create educational materials based on the latest real-life examples. Fundamentally new and very important programs are also emerging — the Faculty of artificial intelligence was opened at Moscow State University this year," he said.
According to him, joint laboratories are also important. At the AI Journey conference, MIPT signed an agreement with AI to create a joint laboratory for embodied AI.
Swordfish Security, a Russian cybersecurity company, is actively working with universities. Since September, two joint educational projects have been launched: the online master's degree program "Modern Technologies of Secure Systems" with UrFU and Netology, as well as the online program "Cybersecurity", developed jointly with HSE and Netology. The basics of AI security are planned to be taught to graduates of universities, cybersecurity engineers, and they plan to launch such a course this year, said Yuri Shabalin, Director of Artificial Intelligence Technology Development at the company.
Izvestia sent a request to the Ministry of Education and Science.
Today, most educational programs are still focused on classical IT areas, while new competencies - working with ML models, data pipelines, and generative AI architectures - are being formed point—by-point and mainly in corporate schools, said a member of the State Duma Committee on Information Policy, Information Technology and Communications, federal coordinator of the Digital Russia party project." Anton Nemkin.
— Companies are competing for experienced developers, increasing salary offers, offering hybrid formats and accelerated development trajectories. However, it is impossible to "grow" specialists quickly: high—quality training takes several years," he stressed.
Many organizations are ready to train employees from scratch, but even here there are limitations — basic digital skills, analytical thinking and an understanding of the subject area in which the model is trained are necessary for effective work, the deputy added. In his opinion, solving the problem requires a systematic approach: expanding specialized programs at universities, developing applied learning based on corporations, stimulating internal academies and mass educational initiatives.
Переведено сервисом «Яндекс Переводчик»
