Russian Federation Council approves law on developing AI technologies
MOSCOW. July 17 (Interfax) - The Federation Council approved a law on regulating relations arising in connection with the development, implementation, use and application of large foundational models of artificial intelligence (AI) at a meeting on Friday.
In accordance with the law, such AI models are understood as a computer program, the main characteristics of which are the performance of a large number of intellectual tasks at a level comparable to that of a person, the provision of information, decision-making or forecasting of results according to goals set by a person, as well as the presence in such a program of at least 1 billion parameters.
The law defines the basic concepts, including the definition of the technology itself, and the principles of regulating relations related to AI, and stipulates the powers of the president, the government, federal executive authorities, other state bodies, state corporations and the Central Bank of Russia in the field of AI application.
The law introduces such sovereign and national categories of artificial intelligence models - for which requirements are established for the localization of data processing on Russian territory with the participation of Russian legal entities, as well as support measures for developers of such artificial intelligence models, including in terms of access to data necessary for training models.
The law provides for granting users the opportunity to place labeling (informational warnings) on information materials in audio and/or visual form created using AI.
Given the importance of the issue of safety of AI application at sensitive facilities, such as state information systems, the law provides for the powers of the government to establish cases in which the use exclusively of sovereign or national large foundational models is permitted.
The law is aimed at creating legal conditions for the accelerated development and implementation of large foundational artificial intelligence models in Russia, ensuring state and technological independence, the security of individuals, society and the state when using AI, as well as increasing the efficiency of public administration and the development of an innovative economy, according to the explanatory note to the document.
Deputy Prime Minister Dmitry Grigorenko previously said that AI models must fully comply with Russian legislation and traditional spiritual and moral values. The document establishes criteria for national and sovereign models, he said. "It was important for us to 'mark' specifically Russian AI in order to provide support and stimulate its implementation in critically important areas," he said.
