Emerging AI-involved crimes are challenging the traditional concepts of Chinese criminal law, with particular reference to actus reus, mens rea, causation and criminal liability. This report examines whether and how traditional criminal law doctrines in China can be applicable to AI crimes. It finds that AI is not yet recognised by the Chinese criminal law as an independent subject of criminal liability, though academic debates regarding this topic are highly intense. Nonetheless, it is safe to say that for the time being the criminal law can still be resorted to in order to deal with most harms generated by weak AI, because AI’s conduct is almost fully reducible to human actors. A key concern is that the criminal law is not sufficient to deal with novel illegal conducts committed by strong or super AI. Thus, it is recommended by Chinese legal scholarship that new AI offences and criminal punishments specifically targeting super AI agents should be introduced by future amendments to the criminal law as an approach to conform with the principle of legality.