Russia’s leaders have been paying close attention to artificial intelligence (AI) technologies for several years now. President Vladimir Putin has said on numerous occasions that the leader in the field of AI would become “the master of the world.” Until recently, however, Russia remained virtually the only large country without its own AI development strategy.
That changed in October 2019, when the country adopted a long-discussed National Strategy for the Development of Artificial Intelligence Through 2030. One of the driving forces behind the strategy was Sberbank president German Gref. The state-owned bank has also developed a road map for developing AI in Russia and coordinated the creation of Russia’s AI development strategy, which is largely corporate, involving the internet giants Yandex and Mail.ru Group, along with Gazprom Neft energy company.
The same companies, along with MTS and the Russian Direct Investment Fund that joined later, have formed a structure known as the AI Russia Alliance, which is tasked with promoting Russia’s AI-based technologies. The alliance was established on November 9, 2019, at the AI Journey conference attended by Putin. The new entity is supposed to lobby for simpler procedures for introducing AI technologies, as well as the participation of business in two areas: developing driverless transportation, and legislation relating to industrial and personal data.
The alliance is expected to coordinate the efforts of the business and scientific communities to attain the objectives set forth in the national AI strategy. It can be concluded, therefore, that the driving force developing Russian AI technologies in the near future will be commercial—and that it will be large IT companies rather than start-ups.
The Russian government finances industry-specific projects—quite generously by Russian standards—which testifies to the high priority assigned to this field. The application of AI in the military industry, in which Russia’s position is traditionally strong, is the subject of particular attention.
Nevertheless, it is unlikely that Russia will become a leader in the development of AI as set forth in the national strategy because of its current lag behind the leading technological powers, as well as some other factors, such as a small venture capital investment market. The successful development of specific areas of AI is more likely, and Russia may be capable of becoming a local leader in those fields.
On the international arena, Russia opposes banning lethal autonomous weapons systems (LAWS) and the military use of AI, but it does engage in dialogue with other states and actors and supports formulating clear universal rules and ethical norms.
Russia’s Priorities
The national strategy stresses two defining time markers for the development of AI in Russia: 2024, by which time Russia is expected to have significantly improved its positions in this field, and 2030, when it should have eliminated its lag behind developed countries and attained global leadership roles in certain AI-related areas. According to the document, Russia’s key AI development priorities include increasing the number of entities involved in technological innovation by 50 percent, and creating а high-performance export-oriented sector equipped with modern technologies in key industries, primarily in manufacturing and agriculture.
The strategy focuses heavily on support for both state and privately sponsored scientific research. Citations of Russian scientific publications in the field of AI, as well as the number of patents and applied technological solutions registered and developed by Russian scientists, are expected to increase significantly by 2024. The government is also working to overhaul existing legal norms to simplify the development and implementation of AI-based technologies (for instance, driverless transportation). The document stresses the need for Russia’s international cooperation on issues of AI-based product standardization and certification.
Another of the strategy’s focus areas is the importance of preparing AI specialists. Apart from recruiting foreign specialists, the document proposes introducing AI-related educational modules at all educational levels, as well as creating respective job development and professional retraining programs. Special emphasis is placed on math and science education and its integration with humanities. According to the strategy, the number of Russian AI specialists is to grow significantly by 2024, and Russian universities are to acquire world-class educational programs that will make it possible to make up for personnel shortages.
The Russian strategy also emphasizes domestic AI-based software development, the creation of AI open-source libraries, better data access and quality, and improved hardware for AI projects.
As for military objectives, the document identifies them purely in terms of “guaranteeing national security.” But the overall strategy clearly favors commercial AI use.
Current Assessments
Unlike China and the United States, Russia isn’t a global leader in AI technologies. Although the strategy states that “the Russian Federation possesses significant potential to become one of the international leaders in the development and use of artificial intelligence technologies,” it’s unlikely the country will be able to achieve this goal in the short and medium term. Nevertheless, certain areas of AI development and application do exist; Russia enjoys strong positions there and can succeed in the future. Russia’s ranking in the AI technology race is hard to establish, since current international ratings use different methodologies. Many of them don’t list Russia at all, which is unsurprising, since the country lacked an official AI development strategy until November 2019.
In terms of individual AI-related technology indicators, Russia has only three supercomputers ranked among the 500 most powerful computers in the world, for example, while China has 228, the United States has 117, and Japan has 29.
At the same time, Russia is doing quite well in terms of computer science education, which includes machine learning. Sixteen Russian universities that offer this major made it onto the list of the world’s 684 best educational institutions, according to the 2019 World University Rankings. However, only two Russian universities—Moscow State University and St. Petersburg’s ITMO (Information Technologies, Mechanics, and Optics)—were in the top one hundred.
The number of AI start-ups in a given country is also a key indicator of its progress in this field. According to TRAXCN statistics, Russia currently has 168 AI start-ups, compared with 6,903 in the United States and 1,013 in China. These estimates appear to be somewhat exaggerated, since the list includes twenty- and thirty-year-old “start-ups,” but still, they make it possible to assess the bigger picture of the state of venture capital investment markets with respect to AI. In part, the small number of Russian AI start-ups can be explained by the domination of established companies in this sphere. This is what distinguishes Russia from most other developed technological powers. According to Russia’s proposed Artificial Intelligence road map, the country has a total of 400 companies working on developing AI.
Military Applications
International ratings obviously don’t include military programs, yet military science and the defense industry have been locomotives for the development of Russian technology for many years. It’s difficult to assess the extent to which AI is used in the military sector due to the classified nature of such projects. Nevertheless, public pronouncements by Russian government officials make it possible to draw certain conclusions. Defense Ministry representatives and other government officials have said repeatedly that the Russian Armed Forces boast a full lineup of AI-based weapons, such as unmanned aerial vehicles, fighter jets, and underwater robots.
In his address to the Federal Assembly on March 1, 2018, Putin stated that Russia had developed an unmanned deep-sea vehicle that is capable of global-range travel and can carry nuclear weapons. These autonomous vehicles, dubbed Poseidon, are slated to join the Russian naval arsenal by 2027.
That same month, Defense Minister Sergei Shoigu urged military personnel and civilian scientists to combine their efforts in working on AI technologies “whose development is necessary to parry possible threats to Russia’s technological and economic security.” Shoigu was speaking at the first “Artificial Intelligence: Problems and Prospective Solutions” conference held at the Defense Ministry. It’s difficult to estimate Russian investments in the military applications of AI, but it is possible to list some structures affiliated with the Defense Ministry that are working on these issues.
The state-owned Rostec corporation has acknowledged that a number of its subsidiaries—the Kalashnikov and Tecmash concerns, High Precision Systems holding, and TsNIITochMash company—were working on developing AI-related weapons. Tecmash, for example, has been incorporating AI into the volley fire systems it produces for several years now. In a December 2019 speech at the Defense Ministry board meeting, Putin said: “Robotic systems and unmanned aerial vehicles are being rigorously introduced and used in combat training, which dramatically boosts the capabilities of armed units and subunits.”
New models of weapons and equipment are being actively tested in Syria. Back in 2016, the Russian Defense Ministry alluded to the use of unmanned aerial vehicles (UAVs) to monitor the ceasefire there. During the next few years, both serial UAV models such as Orlan-10 and Forpost and test models like the Orion medium-altitude long-endurance UAV—potentially capable of carrying weapons—took to Syria’s skies. Autonomous aerial vehicles are one of the most promising areas for the introduction of AI and its military use. The exact number of UAVs in the Russian army arsenal is unknown, but the head of the Russian General Staff’s office for UAV development, Major General Alexander Novikov, estimated it at more than 1,900 units in 2018, and their number has likely increased since that time. In August 2019, the Defense Ministry published a video of the maiden flight of the S-70 Okhotnik-B heavy combat UAV developed by the Sukhoi Design Bureau. The Okhotnik (whose name translates as “hunter”) is likely capable of carrying weapons and striking both air and land targets.
In April 2020, the Defense Ministry announced a closed tender worth up to 387,751,000 rubles ($5.3 million) to complete a scientific project titled “Research on creating an experimental model of neural network development, training, and implementation for the new generation of artificial intelligence military systems” (code name “Kashtan”). In addition, the Era military innovation technopolis is due to be completed in the southern city of Anapa in the fourth quarter of 2020. It will specialize in robotics, AI, information security, and supercomputers.
The military sector is, therefore, one of the strongest in terms of developing Russian AI technologies. Roger McDermott of the Jamestown Foundation believes that in terms of decisionmaking speed, the new AI-based Russian combat control information systems that were showcased during the Tsentr 2019 strategic exercise give the Russian military a strategic advantage over NATO armed forces.
Support Instruments
According to the proposed AI road map prepared by Sberbank under a government contract, Russia is planning to spend 244 billion rubles on the development of AI through 2024. Sberbank itself is planning to invest almost half of that: 112 billion rubles, while another 91 billion rubles will come from the federal budget. The Russian Direct Investment Fund (RDIF) also wants to take part in the venture.
There is also a proposal to create a new 120 billion ruble AI federal project as part of the Digital Economy national program (the road map was also completed under this program, albeit under the Digital Technologies project). The artificial intelligence project was also developed by Sberbank and is supposed to be overseen by the Economic Development Ministry. In other words, there is some confusion now over instruments of AI development in Russia.
In November 2019, the RDIF reported that it had raised $2 billion in domestic and international investments in Russian companies that use AI. The RDIF will also undertake to bring these companies to international markets. According to some media publications, the fund will invest some of its money in creating an Artificial Intelligence Institute at Moscow State University.
Such investment volumes are unique for Russian science. But how significant are they on a global scale? It is certainly far less than the leaders of the tech race spend on AI. China annually invests tens of billions of dollars in AI, including at a regional level. Beijing alone is planning to invest $2.1 billion in an AI industrial park, while Tianjin is considering the creation of a $16 billion AI development fund. In turn, the U.S. government allocated $4.9 billion for AI research in 2020. Venture investors have contributed even more, investing over $8 billion in AI start-ups in 2018 alone.
Nevertheless, the Russian road map budget is quite comparable to what other technologically advanced countries spend on AI development. The UK, for example, revealed plans in 2018 to invest 1 billion pounds ($1.3 billion) in AI development, and Germany is to invest 3 billion euros ($3.3 billion) through 2025. It is assumed that private investors will contribute similar amounts.
The Russian government and Russian technology giants are making significant efforts to bridge the gap between Russia and leading countries when it comes to AI development. In February 2020, Putin endorsed the idea of creating a network of supercomputers and second-tier computing centers in large cities throughout the country.
Russia has already managed to make significant progress in this area. At the end of 2018, the first energy-efficient petaflop Zhores supercomputer was launched at the Skolkovo Institute of Science and Technology (Skoltech). It’s specifically designed to solve machine learning and data-driven modeling problems. In the near future, its capacity is expected to increase to 2–3 petaflop per second, putting it in the same league as the top 500 supercomputers in the world. A year later, Sberbank launched its Christofari supercomputer. Specially designed to work with AI algorithms, the Christofari has an effective performance of about 6.7 petaflops per second, making it the most powerful in Russia and the twenty-ninth most powerful in the world at the time of its launch.
Sberbank and RDIF also intend to invest significant funds in the development of Russian software. In addition, there are state plans to develop a grant system, introduce new educational programs in the field of AI, and include AI training courses in the school curriculum. As a result, the number of Russian publications at top-level AI conferences is expected to increase from 30 in 2018 to 500 in 2030, while the number of AI specialists trained by higher and continuing education programs is expected to grow from 650 to 6,000 over the same period.
Russian businesses are willing to put AI technology into practice, giving Russia a competitive advantage. Microsoft has named Russia the world leader in the active implementation of AI in business. According to its research, 30 percent of Russian companies actively implement AI: the highest number among all countries studied, compared with an average of 22.3 percent.
Russian companies are also actively involved in the development of AI technology. They are, for example, at the forefront of developing driverless vehicles. Although Russia’s position in the Autonomous Vehicles Readiness Index is still fairly weak (it was ranked 22 out of 25 in 2019), Russia has been actively developing this field with the help of a number of companies and organizations including the internet giant Yandex, KAMAZ truck maker, StarLine unmanned vehicle developer, the Central Scientific Research Automobile and Automotive Engines Institute (NAMI), and the Moscow Automobile and Road Construction University (MADI).
Driverless cars have been tested on the roads of Moscow and the Russian republic of Tatarstan since 2019, and in March 2020 the experiment was expanded to eleven more regions. A bill on the use of autonomous vehicles on public roads was introduced to the State Duma in late February 2020. In addition, in March of this year, the Non-Commercial Partnership for Development and Use of Navigation Technologies (NP GLONASS), which includes Yandex, Rostelecom, MTS, and other companies, sent letters to the foreign, transportation, and justice ministries requesting the suspension of certain provisions of the Vienna Convention on Road Traffic that impede the introduction of driverless vehicles. The complete legalization of autonomous vehicles will be a major boost to the development of Russia’s driverless transportation industry.
Russian AI technologies have, therefore, been developing at an accelerated pace, largely due to the greater attention that the Russian government and major tech corporations now pay to this sector. Although state investments and grants can’t adequately replace venture investments, Russia does have an opportunity to significantly improve its positions in the AI race.
International Cooperation
Russia believes it would be premature to discuss limitations on the military use of AI at an international level, citing the lack of commonly accepted terminology and the absence of lethal autonomous weapons systems (LAWS) as reasons for its position. In part, this is because Russia views itself as one of the leaders in developing AI and robotics. Many countries that have embarked on successful LAWS projects—the United States, Israel, the UK, and others—hold similar positions.
Today, the main forum regulating AI is the UN. Since 2017, a specially created group of government experts (GGE) has been discussing LAWS issues in the context of the Inhumane Weapons Convention. In the past, similar discussion formats have led to the banning of cluster bombs, napalm, blinding laser weapons, and anti-personnel mines. Russia is represented in the GGE by officials from the foreign, defense, and economic development ministries.
Thirty countries currently insist on restrictions on the development and use of LAWS. Russia’s official position is that “work on definition and basic functions should mainly be guided by the ultimate goal of LAWS discussion—i.e., explore possibilities to use this weapon type in a most adequate manner in future and maintain a due level of human control over it. At the same time, specific forms and methods of such control should remain at the discretion of states.” Russian officials underscore that the work of the GGE should not “upset the balance between humanitarian concerns and the interests of defense security of the participating states.”
Another international platform for discussing AI problematics is UNESCO, which has had an ad hoc expert group working since March 2020 on formulating global standards and recommendations on AI ethics. Russia is represented in the group by Professor Maxim Fedorov, director of the Center for Computational and Data-Intensive Science and Engineering at Skoltech. The group plans to meet through August 2020, by which time it is expected to have come up with internationally applicable recommendations on formulating ethical principles for AI development and use.
The Committee on AI under the Commission of the Russian Federation for UNESCO is called upon to consolidate the Russian position on this issue. That committee is chaired by Alexander Kuleshov, the head of Skoltech and a member of the Russian Academy of Sciences. Its first session featured a discussion on differences in global understanding of AI ethics, at which Fedorov said: “We need to move from polarity to international cooperation.”
On the whole, Russia seems to be supportive of developing international cooperation in the field of AI and formulating universal and legal norms in this sphere. As for Russia’s position on the possible LAWS ban, it proceeds from efforts to lobby the interests of the domestic defense industry, and is in line with the behavior of most countries that produce unmanned vehicles.
As of today, Russia cannot be described as a leader in the AI race. Even if AI development becomes Russia’s highest priority, Moscow essentially has no chance of catching up with Washington and Beijing in this field. Nevertheless, smart investing and capitalizing on the country’s competitive advantages, such as highly skilled computer specialists and large domestic IT companies, could deliver results in the medium and long term. Russian positions have been traditionally strong in the military sector, but the successful development of other AI-related sectors is also possible. Under favorable conditions, Russia is quite capable of becoming a serious player and even a local leader in certain areas.