Abstract
The article is devoted to the analysis of the legal nature and regulation of artificial intelligence in the context of rapid digital development. The paper compares approaches to legal regulation of AI in Kazakhstan, the USA, the European Union and China, identifies their fundamental differences and points of intersection. Special attention is paid to the draft Digital Code of the Republic of Kazakhstan and the Concept of Artificial Intelligence Development for 2024–2029, which reflect an attempt to build a holistic legal model combining ethical norms, technical standards and mechanisms of legal accountability. The article reveals differences in regulatory philosophy reflected in national strategies, regulations, and ethical declarations. The conclusion is drawn about the need for flexible, adaptive and multi-layered legal regulation that can take into account both the technical characteristics of AI systems and the risks associated with their autonomy and impact on fundamental rights. The results of the study indicate the importance of moving from declarative norms to operational mechanisms, including the legal status ofAI, certification of algorithms, ethical audit, transparency of decisions and allocation of responsibility.