Vladimir Putin approved the updated national strategy for the development of AI until 2030

Russian President Vladimir Putin approved an updated national strategy for the development of artificial intelligence (AI) until 2030. It sets out the main indicators, for example, an increase in the readiness of key industries for its implementation to 95% compared to 12% in 2022 and an increase in annual costs for this to 850 billion rubles. The strategy also included a reminder of the tense foreign policy situation: the challenges now include restrictive actions on the part of unfriendly countries.

Decree of the President of the Russian Federation dated February 15 No. 124, published on the portal of legal acts, makes significant changes to the national strategy for the development of AI until 2030. The first version of the strategy was approved in 2019. The updated document sets new goals for the strategy. Thus, by 2030, the share of workers with skills in working with AI should be 80% versus 5% in 2022, and the annual number of graduates of educational organizations who have mastered programs in the field of artificial intelligence should exceed 15 thousand (in 2022 there were just over 3 thousand). And finally, by 2030, 80% of Russians will have to “trust AI technology,” while in 2022 this figure was only 55% (how the level of trust is measured is not clear from the document).

The updated strategy also includes the definition of large generative models and large fundamental models. We are talking about AI solutions that can interpret information based on a request and generate texts, videos, images, etc. (These are, for example, ChatGPT, GigaChat.) It is noteworthy that the authors of the strategy, on the one hand, recognize the high potential of generative AI , stating that its use can annually increase global GDP by 1–2%, but on the other hand, they see risks in its use. “AI models create an image in a second based on any text description or sketch, which creates the threat of disseminating prohibited information, violating copyrights and generating erroneous information,” the document says.

The volume of expenses of organizations on the use of AI by 2030 should be at least 850 billion rubles. per year compared to 123 billion rubles. in 2022. The total GDP growth due to the use of AI technologies in 2030 should be at least 11.2 trillion rubles. compared to 0.2 trillion in 2022.

The new version of the strategy also describes the difficulties that the authorities will have to overcome in its implementation. These are a lack of computing power, personnel shortages, new types of threats in the field of information security, the emergence of new types of threats in the field of AI development that are “not typical for other areas of information technology,” as well as “limited access to artificial intelligence technologies due to unfair competition with sides of unfriendly foreign states.”

Russian President Vladimir Putin announced the signing of a new version of the national strategy at the plenary session of the AI ​​Journey conference on November 24, 2023. “Russia should become one of the most comfortable jurisdictions for the development of AI,” he noted. To achieve this goal, the February 15 strategy text sets out, for example, the types of data that AI developers should have access to. This is medical data, anonymized personal data, industrial data, as well as from the field of science, education and culture, etc.

To ensure the availability of big data needed for AI training, the government late last year approved the creation of a state technology development platform. Through it, developers will be able to access computing infrastructure and data sets (see “Kommersant” dated October 18, 2023).

The areas where the implementation of technology is still minimal include, first of all, law enforcement activities, reminds Igor Bederov, head of the investigation department at T.Hunter: “It is considered lagging behind in terms of technology around the world.” In medicine, AI technologies are also not being developed actively enough, although they are extremely necessary and effective. Education can also be included in these areas, the expert believes. “AI has great potential in treating patients, but so far its application has been limited, because training neural networks requires the same type of labeled data, and so far there are very few such data sets,” agrees Sergei Lezhnin, CEO of the Doctis group of companies (telemedicine service).

Tatiana Isakova