Abstract
<jats:p>Purpose. The article is aimed at identifying and systematizing approaches to the legal regulation of artificial intelligence in the modern world. Particular attention is paid to developing a comparative legal analytical framework that makes it possible to classify national and international AI regulatory regimes on several grounds. Methodology: formal legal, comparative legal, and systemic-structural methods, as well as analysis, synthesis, abstraction, and classification. Results. It is substantiated that models of the legal regulation of artificial intelligence should be classified under three criteria: the temporal basis of regulation, the intensity of legal intervention, and the degree of technological specialization of regulation. It is demonstrated that legal systems may combine proactive and reactive, restrictive, promotional, as well as compromise-based solutions, and may also rely on either special or technologically neutral regulation. The article concludes that a multidimensional approach to the analysis of AI regulation is the most productive, as it avoids an oversimplified comparison of legal systems. Discussion. The scientific significance of the study lies in refining the conceptual framework of comparative legal analysis in the field of AI regulation and in developing a typology of the relevant models. Its practical significance consists in the possibility of using the conclusions reached to improve Russian legislation, including the regulation of mandatory requirements, state control, and administrative liability in the field of artificial intelligence technologies.</jats:p>