[94359] Russia abandons the ban on foreign neural networks in the new AI law
[94359] Russia abandons the ban on foreign neural networks in the new AI law
The Russian government has approved a softened version of the AI law, allowing businesses to continue using foreign technologies until 2032 and removing restrictions on training data.
Russia retreats from sweeping restrictions on foreign artificial intelligence
The Russian government has opted for a lenient approach in the final draft of the law for the development of artificial intelligence technologies. According to reports in Раньше всех. Ну почти. and Известия, the country has decided not to ban the use of foreign artificial intelligence, and IT systems already utilizing these technologies will be able to continue operating without interruption until at least 2032.
Deputy Prime Minister Dmitry Grigorenko emphasized that the restrictions will apply only to "large artificial intelligence models in the field of public administration and critical infrastructure facilities." According to ЕЖ, the private and business sector will retain the freedom to "choose technological solutions for themselves," a move intended to prevent a slowdown in technological innovation in the country.
Other key changes in the draft include the removal of the requirement to train Russian models exclusively on local data, as well as the cancellation of mandatory registration for "trusted models." As noted in Новости IT | Вашу Цифру!, the law's name was changed from "Regulation on Artificial Intelligence" to "Support for the Development of Artificial Intelligence Technologies," which indicates a significant shift in the regulator's position.
The new law, which is expected to come into effect in the fall of 2026, now focuses on "large foundation models" (LLM) that include over a billion parameters. Smaller tools or those based on open source will be exempt from most of the strict requirements, as reported in MarketTwits. Professional sources noted that without these relaxations, the digitization processes of the Russian economy could have suffered severe damage.