The United States and China compete for the domination of AI while the model of Russia is lagging behind

• Russia presented its first LLM, Gigachat Max, as part of an AI national strategy.
• But the model is “commonplace” and is late on us and the Chinese offers, IA experts told Bi.
• While the war in Ukraine has slowed down development, Moscow could still develop a military AI.

Russian President Vladimir Putin wants his country to participate in the world race to build AI, beating models out of China and the United States. But its flagship model of great language, or LLM, is not even the best to speak Russian.

On the Russian language version of LLM Arena – where users will compare and classify the responses of different LLM – Gigachat Max comes joint at the time of writing, behind various versions of Claude, Deepseek and Chatgpt.

Yandexgpt 4 Pro, an LLM developed by the Russian search engine Yandex, is even lower, with 18th joint.

On the English version, none appears in the ranking of more than 170 llms.

Gigachat Max was developed by Sberbank, belonging to the State of the State of Russia. When its last iteration was launched in November, its main developer based in Moscow, Evgeny Kosarev, said on Linkedin that he was “close to GPT4O in quality on Russian and English”.

But the experts told Business Insider that, despite Putin who emphasized the development of AI as a crucial avenue for the Russian foreign policy, Gigachat Max has months behind American and Chinese competitors. The country’s war against Ukraine also drained him with expertise.

The spokesperson for Gigachat Max and Yandex did not respond to the request for comments from Business Insider.

A “banal” model

For the moment, Gigachat Max, the most developed LLM in Russia, is “Better,” said Bi Bi.

On “references” – the standardized tests for the efficiency of the AI ​​- the scores of the models “are much lower”, he said, adding that they do not exceed any of the advanced or “border” models, and do not imply any particular innovation.

Ben Dubow, main member of the Center for European Policy Analysis and CTO of the Omelas data analysis company, added that Gigachat Max did not have an advantage in many ways.

Although he manages mathematics well, in the Russian language, he is far behind most Western and Chinese LLM on certain landmarks, Dubow wrote in Moscow Times in January.

He said the main LLM developed in the United States had a year ahead of the current level of Gigachat Max on the “massive understanding of multitasking language” of industry, or Mmlu, which tests the general knowledge of an LLM and the problem of solving problems in textual responses on a wide range of subjects.

Dubow also told Bi that most AIs were held to more advanced references, Mmlu “almost considered as past at this stage”.

“The decline in American and Chinese models on Russian language prompts is an absolute priority for the Russian government’s strategy, but Max has not succeeded,” said Dubow.

Related stories

The war in Ukraine retains the development of Russian AI

Russian President Vladimir Putin repeatedly stressed the importance of AI, including at a December conference where he praised Gigachat Max and said Russia was ready to help other nations develop an AI.

Samuel Bendett, specialist in Russian military technology at the Center for Strategic and International Studies, told Bi that AI was “a status” for Russia.

But according to a global AI ranking produced by the startup of the British media Tortoise Media, Russia is the only one of the five “great power” countries – the United States, China, France, the United Kingdom and Russia – not at the top of the list. Russia is classified 31st.

Bendett has appointed several factors holding the Moscow AI sector.

The private sector of Russia is too small to promote real competition, with almost everything that is supported by the government, He said.

Although Sberbank is getting more and more as a technological business, “there is no equivalent to Openai and Microsoft or Google or Huawei or Alibaba,” he continued.

In addition, the invasion of Ukraine by Russia has isolated it from both global expertise and collaboration, as well as access to technology as well as the microchips necessary to train and effectively manage complex AI models.

“The history of the Russian AI industry is, in many ways, Putin’s expansionism undergoing the global position of Russia,” said Dubow.

2014 – When Russia annexed Crimea – was a transformative year for AI in the West and in China.

Meanwhile, 2022, the year when Russia launched its large -scale invasion of Ukraine, was launched by the Chatgpt, causing generation of generator.

THE War in Ukraine Accelerated a large brain leak from Russia, according to Dubow.

Bendett added that Russia does not have “hundreds of thousands” of high -tech researchers, although he said that he believed that many “technological refugees” who left Russia to prevent the project began to fall back.

Putin recognized the problems last year, blaming the “hostile countries” for roadblocks and promising to increase the number of graduates in AI technology to more than 15,000 per year by 2030, reported the Russian TASS news agency, citing government documents.

The report indicates that 3,000 graduated in 2022.

In comparison, the United States had more than 73,000 graduates in fields related to AI in 2023, the majority of which were international talents, according to the Center for Security and Emerging Technology.

Serhii Kupriienko, CEO of Swarmer, a Ukrainian startup specializing in AI systems, told BI that during the next decade, the American and Chinese LLM will help them to evolve their “exponentially” economies by increasing productivity in various sectors, by creating jobs in AI and accelerating in an exponential way.

Meanwhile, Russia’s difficulties with AI mean that its most likely path is to “be subordinate to China and rely on what China’s production,” said Dubow.

The “Saint Grail for AI” could stimulate the military of Russia

Repeated public statements of the Kremlin on the AI ​​and the current war in Ukraine led certain analysts to conclude that Russia could secretly develop an LLM with double use with military applications.

In 2022, a Russian official announced the creation of a department for the development of AI within the Ministry of Defense.

“Russia is considering AI as a transformer tool for its soldiers,” wrote military analysts from the Saratoga Foundation Timothy Thomas and Glen Howard in a February review of Russian writings on military AI.

Vitaliy Goncharuk, who chaired the Ukraine AI Committee between 2019 and 2022, believes that Russia could form its AI on large quantities of data on the battlefield generated in Ukraine.

Posts and telegram channels, drone sequences, satellite imaging, sound sensors, civil reports and hacked equipment of the Ukraine Delta -based management system – which feeds Ukrainian commanders with data on the battlefield – all provide enough equipment, said Goncharuk.

The AI ​​has developed on this subject would not only help Russia improve its precision in the identification of objectives, but also to help it plan its decision -making and its first line operations, said Goncharuk. This could even predict the future decision-making of Ukraine and future battlefield operations, he added.

Ukraine has also gathered large amounts of data on the battlefield of three years of war – something which is “really the holy grail of the formation of your models and systems of AI on the recognition and selection of the target of the battlefield”, Bendett said bi.

It would be difficult to imagine that Russia is not silently using this data, he added.

“They are constantly allowing allusions to this,” he said.