Measures and dangers for the development of artificial intelligence were discussed at Finopolis

Artificial intelligence (AI) technologies in Russia cannot do without the help of the state, as follows from statements made on Wednesday at the plenary session of the Finopolis forum – but bankers consider it important not to regulate this area. Thus, according to the head of Sberbank German Gref, a strict approach to regulatory issues actually killed the development of AI in Europe. The Central Bank, which oversees the development of technology in the Russian financial sector, is not going to attack AI “with fire and sword”, proposing to use a risk-based approach to its regulation, while banks hope that the development of AI will also receive financial government support – for example, in the form of government orders.

Yesterday at the plenary session of the Finopolis forum “Artificial intelligence: the path to the common good or the road to slavery?” The head of Sberbank, German Gref, said that for the Russian Federation to lead in this technology, the state must first of all not regulate it. “A lot depends on the state, especially if the state begins to regulate, so the first and most important thing is to try not to regulate this area, this is the most that the state can do,” he said, noting the reasonable concept of the Bank of Russia, set out in a recent published advisory report on the use of AI in the financial market and based on a risk-based approach to the use of technology.

German Gref referred in his speech to the examples of the USA and China, where all conditions are created for the development of AI without “overregulation”, as well as to the experience of Europe, where a strict approach to regulation actually killed the development of AI. “Large companies are trying to relocate somewhere outside the jurisdictions of the European Union. It is very important for us not to repeat this,” the banker insisted. According to him, the publications of generative models by Sberbank and Yandex have already attracted the attention of the prosecutor’s office – the reason was that Sberbank’s neural network added the domes of St. Basil’s Cathedral to the Russian flag, and deputies considered this “a mockery of our national flags”; due to the resulting fine-tuning of the AI, Sber’s model lost 12% in creativity and accuracy. “It’s good that we came across such a modern prosecutor, with a sense of humor, studied everything and, thank God, we weren’t brought in anywhere. He says: “Guys, come on, go to work, but be careful,” said Mr. Gref.

The head of VTB, Andrey Kostin, also spoke about the protection and support of the state. “Measures of indirect support, measures in the form of orders for key things, I think, need to be done,” he said, explaining: “Giant amounts of money are needed for the development of technology, we (the banks. – Kommersant), of course, also need to unite, in This can also be the role of the state.” However, Mr. Kostin immediately asked the head of the Bank of Russia, Elvira Nabiullina, for regulatory relief for the IT investments of the financial sector, citing the fact that VTB’s budget for the next year is 160 billion rubles, a significant part of which is expenses on AI: Sberbank is already making money on AI spends more than it spends, but VTB has not yet reached this stage, “we invest and invest.” “These investments are considered intangible assets and must be deducted from regulatory capital. Even the European supervisory authority, which is especially known as the “strangler” of the banking sector in Europe, even offers installment plans for three years. We cannot yet get the Central Bank to make progress on this topic,” the banker complained.

However, basic answers to bankers’ concerns were given even before they were heard – in a statement by Elvira Nabiullina: although the development of AI is associated with risks, the Central Bank is not going to “go with fire and sword against artificial intelligence”, but will focus on finding a “golden mean” “to create incentives for the development of this technology, but also to protect citizens.” Intending to “use the so-called risk-based approach,” the head of the Central Bank believes that the market will cope with some of the risks, especially at the first stage, through self-regulation. Strict legislative regulation of certain aspects may be required – “but it should be introduced only after a thorough risk analysis and after consultation with the market.” “I think we will discuss all this with the government,” she noted.

Venera Petrova