On December 8, according to the Wall Street Journal, Russian President Vladimir Putin has repeatedly stated that Russia must lead the world in the field of AI. But the reality is that while other countries are chasing after each other, Russia has become a bystander. Just as the United States and China are competing for dominance in AI models and applications, and countries in Europe and the Middle East continue to invest resources in building computing infrastructure, the war in Ukraine has derailed Russia's once ambitious AI ambitions.


russian president putin

Model lags significantly behind

On the Russian version of the AI ​​model evaluation platform LM Arena, the top Russian model ranked only 25th, even behind the early versions of ChatGPT and Google Gemini. According to the "Global AI Vitality Tool" released by Stanford University in November (which measures the strength of each country's AI ecosystem), Russia ranks 28th among 36 countries.

Western sanctions have cut off Russia's access to critical hardware such as computer chips and restricted its domestic production capabilities. Today, Russian companies have to rely on middlemen in third countries for everything from high-end chips to simple ChatGPT subscription services. Moscow also relies heavily on China.

Compounding the problem is the brain drain problem. Since the outbreak of the Russian-Ukrainian conflict, top talent has continued to flee Russia. Isolated from the international market, Russian AI companies received only about $30 million in venture capital last year. In comparison, OpenAI alone raised more than $6 billion last year.

“Russia has been lagging behind for many years in developing its own AI,” said Yury Podorozhnyy, a former Russian technology executive.


Sanctions have made it difficult for Russia to obtain critical hardware

Bodorozhny has witnessed the development of AI in Russia firsthand. Over the years, he has been committed to developing Russian versions of Google Maps and Netflix, and has also participated in the development of machine learning tools that are now the core of the AI ​​wave. In 2022, shortly after the Russia-Ukraine war broke out, he boarded a plane and left Russia with his pregnant wife.

Already lost

“Russia has lost this competition and can no longer catch up,” said Podorozhny, who now lives in London and serves as the chief AI officer of Finom, a financial technology startup.

A Moscow-based AI company executive agreed with Bodorozhny’s judgment. He pointed out that Russia's economic and geopolitical isolation makes it difficult for companies to obtain financing and to break through the relatively limited domestic market to achieve large-scale development.

As the potential of AI to reshape the global economy becomes increasingly apparent, countries are scrambling to strengthen control over their own AI infrastructure, data and models to avoid strategic dependence. The same is true in the military field. From battlefield decision support tools to automatic defense systems, the combat readiness of various countries increasingly depends on the strength of sovereign AI.

For Moscow, this strategic task is particularly urgent given the escalating confrontation between Russia and the West.

“We must not allow dependence on foreign systems in critical areas,” Putin stressed at an AI conference last month. “For Russia, this is about national, technological and values ​​sovereignty.”

Russian officials have acknowledged the shortcomings but said domestic models rival foreign products and are improving rapidly. However, some people put it more bluntly.

“Most of our industries are still very, very far away from actually using AI.” Herman Gref, CEO of Russia’s state-owned Sberbank, said earlier this year that the bank is taking the lead in promoting AI development in Russia.


Russian humanoid robot falls

It’s not just Russia’s AI models that are lagging behind. In November this year, at a technology conference held in Moscow, Russia’s first AI humanoid robot AIDOL hobbled onto the stage accompanied by the theme song of the movie “Rocky”. It tried to wave, but then fell over. The organizers immediately interrupted the demonstration and moved the robot away from the site. Organizers say the robot will "learn from the consequences of its actions."

Severe core shortage

In fact, long before the Russia-Ukraine conflict broke out, Russia's chip design had mainly relied on foreign technology, and local chip production capacity was limited. Some advanced Russian-designed chips were manufactured by TSMC.

In 2022, the United States announced a ban on the sale of high-tech products, including semiconductors, to Russia. The ban also extended to certain foreign products produced using US equipment, software or design drawings. South Korea and Taiwan, which dominate the field of high-end chips, and Japan, which is strong in chip manufacturing materials and tools, also quickly implemented export bans on such products. TSMC immediately stopped exporting semiconductors to Russia.

The move leaves Russia suddenly unable to directly purchase high-performance graphics processing units (GPUs) critical to training AI models, including Nvidia's latest chips. A Wall Street Journal analysis of United Nations trade data shows that last year, the number of GPUs and other computer chips necessary for AI development dropped by Russia by 84% compared with pre-war levels.

"Russia has a serious shortage of GPUs, and the existing stock cannot even meet current demand," noted Podorozhny, who still maintains contact with Russian AI experts. The above-mentioned AI executive in Moscow agreed with this and said that advanced GPUs can only be obtained through "informal channels" such as intermediaries.


GPUs and other computer chips imported from Russia

Although analysts believe that it is possible for Russia to purchase chips through its Central Asian neighbors, Bodorozhny emphasized: "The key problem is that it is difficult to achieve large-scale procurement."

Sanctions also cut off Russia's supply of materials and components needed to reestablish local production of such products. Currently, Russian chipmakers aim to produce 28nm process chips in domestic factories by 2030. U.S. chip manufacturers are currently beginning to transition to 2-nanometer process chips.

Since Russian bank cards can no longer be used outside the country, it is now difficult for Russians to even pay for foreign models such as ChatGPT. Russian websites are rife with workarounds, from using cards issued in Kazakhstan, Armenia or the United Arab Emirates, to buying online gift cards or working through payment intermediaries. Telegram channels are full of users communicating tips and middlemen selling information on their services.

Sanctions have made Russia's AI development deeply dependent on China, which is exactly the same as the dependence of Russia's overall economy on China. Earlier this year, Putin ordered the Russian government and Federal Reserve Bank to cooperate with China on AI research and development, deepening this trend that has continued to strengthen since the Russia-Ukraine conflict. At the same time, some of Russia’s leading models are also built on Chinese open source models.

brain drain

Even if Russian companies can find workarounds to import technology or hardware, the country continues to lose another key resource: talent.

Russian official data shows that at least 100,000 IT experts left the country in 2022 alone and have not returned yet. The Russian Ministry of Labor predicts that Russia will face a shortage of more than 400,000 IT talents by 2030. Some analysts believe the actual number may be higher.

Anna Fedosova, who has worked in Russia's tech industry, estimates that 70% to 80% of the country's top AI talent has left Russia. Fedosova currently lives in Prague. She is the co-founder of Copilotim, a human resources automation compliance platform. The startup uses AI technology to automate human resources management and compliance processes.

"It is currently very difficult to recruit top AI engineers in Russia," Fedosova pointed out, "This makes it challenging to achieve AI breakthroughs in Russia. This is undoubtedly a major loss for the country."