ISSEK experts assessed the practice of introducing artificial intelligence technologies in Russia

Despite the growing popularity of solutions based on artificial intelligence (AI), 60% of organizations implementing this technology spend less than 1% of their digitalization costs on this process, experts from the HSE Institute for Statistical Research and Economics of Knowledge (ISSEK) found. Among the most popular AI technologies are computer vision, speech recognition, recommendation systems, biometrics and natural language processing. According to a survey of companies, most of all they need AI to improve or create products and business processes.

HSE ISSEK experts, as part of monitoring the development of artificial intelligence in the Russian Federation, examined 2.3 thousand organizations that use this technology from 36 regions. It turned out that two-thirds (65%) of respondents are still using AI in experimental mode, three-quarters are using it together with other digital technologies (such as industrial software, the Internet of things, communication services).

The top five most popular AI technologies include computer vision (78.7% of respondents), speech recognition and synthesis (62%), recommendation systems based on predictive analytics and big data (40.7%), biometrics (37.5%) and natural language processing (28.2%). According to the director of the ISSEK Center for Strategic Analytics and Big Data, Konstantin Vishnevsky, “after the first wave of visionary perception and high expectations for AI technologies, solutions based on them are gradually becoming an integral business tool.”

The main barriers to the spread of AI are the significant amount of investment required and the lack of personnel. More than 60% of organizations spend less than 1% of their digital technology adoption and use dollars on AI. Experts attribute this wariness to the fact that AI projects usually take a long time to implement and have a delayed effect, which is also difficult to assess in advance due to the lack of expertise and standards for implementing solutions and the uniqueness of projects for each individual company.

General Director of the Agency for Transformation and Economic Development Vladislav Onishchenko notes the lack of well-developed economic models for the implementation of AI, supported by a large amount of available data. At the same time, according to him, at the initial stage, significant investments are required in the development of technology and the restructuring of business processes.

ISSEK monitoring showed what results companies expect from AI-based applications. More than half (54.9%) of respondents expect to improve existing products, 37.5% – to optimize business processes, almost a third (31.6%) – to create a new product for the organization, about the same number (30.8%) – to create new business process. The main mechanism for creating AI-related improvements is research and development: more than 60% of the organizations surveyed conduct it. More than half of the respondents (52.5%) implement AI innovations by creating software and databases, almost 40% implement marketing and promotion tools, and about 17% of organizations carry out engineering work.

It was previously reported that the average level of implementation of AI solutions in 2021 was 21% (with a global figure of 54% in 2020). The authorities intend to increase this figure to 50% in 2024. According to expert calculations cited by Dmitry Chernyshenko, the economic effect for organizations from reducing operating costs when using AI technologies by the end of 2023 could amount to 400 billion rubles, and by 2025 it could grow to 1 trillion rubles. The authorities expect to expand the spread of AI, including by linking the use of technology with access to government support, in particular, in agriculture, industry, transport and tourism (see Kommersant on June 19).

Venera Petrova