CAN ARTIFICIAL INTELLIGENCE BE A SUBJECT OF CRIME?

Аннотация

This article explores the possibility of holding artificial intelligence (AI) criminally liable and recognizing it as a legal subject. The discussion on this issue has intensified, particularly after the granting of citizenship to the robot Sophia by Saudi Arabia. Legal scholars express differing views regarding the recognition of AI as an independent legal subject. On one hand, N. Nevyans opposes granting legal subject status to AI, as this would imply the application of human rights standards to artificial intelligence. On the other hand, F. V. Uzhov argues that AI is capable of forming autonomous behavior and therefore requires specific legal regulation. The article examines three possible approaches to recognizing AI as a legal subject: as a legal subject analogous to a natural person; as a legal subject analogous to a legal entity; as a new, limited legal subject. Under current legislation, AI cannot be recognized as a subject of criminal liability, as criminal law considers only natural persons as subjects of crime. Nevertheless, the rapid development of AI necessitates the adoption of new regulations in the fields of civil law, intellectual property, taxation, and e-commerce. The author suggests that recognizing AI as a limited legal subject — through the application of legal fiction — could offer a viable solution to this issue. Such an approach would make it possible to assign liability for harm and negative consequences caused by autonomous AI systems and would stimulate the development of new institutions within criminal law.