Conceptual issues of legal regulation of artificial intelligence in the Russian Federation in the context of the adoption of a framework law on artificial intelligence: security and development

The subject of the research is the conceptual foundations of special legal regulation of artificial intelligence in the Russian Federation. The author proposes a concept for a draft law on the fundamentals of regulating artificial intelligence in the Russian Federation. The developed draft law suggests several amendments to the current version of the draft published by the Ministry of Digital Development of the Russian Federation. Some elements of the published draft are proposed to be revised (for example, adding the developer of the artificial intelligence system to the list of subjects). Special attention is paid to the provisions of the developed draft federal law that allow for internal coherence of the regulation. Primarily, this refers to the foundations of risk-oriented classification of AI technologies, the proposed special legal regime for socially significant communications, mandatory labeling of synthesized and substantially modified content, the regime of a verified information environment, and the distribution of responsibilities among participants in the technology's life cycle. The empirical basis of the research comprises theoretical constructs and practical examples from the spheres of healthcare, education, finance, public administration, and critical information infrastructure. Through the analysis of cases, systematic interpretation of norms, and legal forecasting, the insufficiency of general legislation and the need for a framework law with preventive mechanisms have been demonstrated. The scientific novelty lies in justifying the model of a framework federal law on artificial intelligence, where central importance is attached not only to general definitions and classification of systems by risk categories but also to special regimes for the circulation of artificially created (substantially modified) content. It is argued that three interrelated institutions may be particularly significant for Russian legislation: the regime of socially significant communications, mandatory labeling of synthesized and substantially modified content, and the regime of a verified information environment. It is shown that preventing harm associated with the use of artificial intelligence should become one of the main guiding principles of the future law. It is concluded that the development of legislation in the examined area should be based on the adoption of a framework federal law that combines regulatory and preventive functions.