Research On The Criminal Responsibility Of Artificial Intelligence

Since the concept of artificial intelligence appeared in the 1950 s,its development has undergone many changes.In a narrow sense,artificial intelligence refers to a kind of science and technology based on human settings and requirements,which can simulate and extend human intelligence through certain theories,methods and application systems.Generalized AI also includes all kinds of intelligent carriers carrying the above technologies,namely,artificial intelligence agents.In recent years,artificial intelligence has gradually evolved from a technical concept to a social concept,and the legal issues related to artificial intelligence emerge in endlessly,and gradually turn from the civil field to the criminal field,which brings many challenges to the criminal rule of law in China.The concept of AI crime should be discussed according to the stages of AI development.In the weak AI stage,the AI crime refers to the behavior that the actor uses the AI technology and related carriers to implement,which has certain social harmfulness and criminal illegality,and should be punished;in the strong AI stage,the AI crime also refers to the behavior that the AI technology and related carriers,which are out of human control,are implemented based on independent will and independent consciousnessAcertain social harmfulness and criminal illegality,which should be punished by penalty.At present,the crime involving artificial intelligence is generally regarded as the criminal tool of the perpetrator,and the subject of criminal responsibility is still the natural person or unit that uses artificial intelligence to commit the crime.It is of great significance to study how the criminal liability of the possible future AI independent of the control of the designer,maintainer and user.On the theoretical level,criminal responsibility refers to the responsibility that the actor should bear because of his criminal act,and the judicial organ on behalf of the state makes negative evaluation and condemns the actor according to the criminal law.Generally speaking,the assumption of criminal responsibility needs to consider four aspects: subject,crime,behavior and penalty.Affirming that AI can bear criminal responsibility independently is bound to break through the current theory of criminal responsibility in criminal law,including the breakthrough of subject level,the breakthrough of fault level,the breakthrough of behavior level and the breakthrough of penalty level.At the level of legislation and practice,although there is no strong artificial intelligence with independent will,on the basis of scientific analysis and prediction of the possibility of strong artificial intelligence,prospective research can avoid the rule vacuum in the future criminal law,and prevent the situation that the crime of artificial intelligence cannot be relied on or cannot achieve the unity of crime and punishment.For the crimes of artificial intelligence in practice,we must strictly abide by the provisions of the current criminal law before we complete the legislative confirmation of the identity of the artificial intelligence subject.At the same time,we should adhere to the modesty of criminal law and not use criminal law too much to interfere with the positive development of artificial intelligence.In the future legislative direction,the criminal law should keep pace with the development of the society.The strong AI crime that may appear in the future has serious social harm and can't be treated according to the current criminal law,so we should make forward-looking legislation,add new crimes according to the characteristics of AI crime,and add penalty methods applicable to AI.