Introduction. In recent years, the question of whether highly developed robots with artificial intelligence elements can be recognized as subjects of law, that is, have rights and obligations in the legal sense, has been actively discussed in Russia and abroad. Theoretical analysis. The study was conducted, and the arguments for and against recognizing robots with artificial intelligence as subjects of law were presented. Empirical analysis. The author carried out an analytical review of the legislation of the Russian Federation, the European Union, the USA and China regulating the legal status of robots with AI, and identified similarities and differences in their approaches. Results. In no country in the world AI robots have legal personality. All examples of granting them this status are either symbolic in nature, or remain proposals and discussions without legislative consolidation. The dominant world practice proceeds from the fact that a robot is an object, a mechanism belonging to a person or an organization, and in relation to it the rights and obligations belong to its owners, developers and users. Both in theory and at the level of political decisions, it is noted that the technical and social maturity of AI is insufficient to consider it an independent entity. An AI robot lacks knowledge and moral responsibility, and equating it with a human would be temporary and risky. Such an electronic legal personality may weaken the real responsibility of people, becoming a way of avoiding corporate responsibility, but all this may change in the future.