| According to the “smart” degree of intelligent robots,we can divide the artificial intelligence era into three stages-the era of ordinary intelligent robots,the era of weak artificial intelligence and the era of strong artificial intelligence.Ordinary intelligent robots,weak intelligent robots,and strong intelligent robots can all achieve the replacement of human brain functions to a certain extent.Among them,the difference between ordinary intelligent robots and weak intelligent robots is whether they have deep learning ability.The difference between weak intelligent robots and strong intelligent robots is whether they have independent identification and control capabilities,that is,whether they can be implemented under the control of autonomy and will.In short,the history of evolution from ordinary intelligent robots to weak intelligent robots to strong intelligent robots is actually enhancement of the recognition and control capabilities of a robot,and the effect of human consciousness and will on “behavior” is gradually weakened.With the continuous evolution of intelligent robots,there is a trade-off between human and intelligent robots in their ability to control and determine “behavior.”The development of artificial intelligence technology has gone through the era of ordinary intelligent robots,is experiencing the era of weak artificial intelligence,and will eventually usher in the era of strong artificial intelligence.As Hawking said,“We are at the entrance to a beautiful new world,and this is an exciting world full of uncertainty.” While artificial intelligence technology promotes economic development,improves people’s living standards,and brings various “surprise” to human society,it also causes many risks and uncertainties.On the one hand,it promotes technological progress,encourages technological innovation,on the other hand,it maintains the bottom line without serious risks.We should actively arranges,and strives to explore the criminal law system of artificial intelligence crimes in the era of artificial intelligence,and strives to prevent and control the technological risks in the era of artificial intelligence.It is the mission of criminal law and criminal law scholars in the era of artificial intelligence to give full play to the maximum value of artificial intelligence technology,to make plans for the stable and healthy development of society,and to escort the harmonious and safe development of society.According to the content layout,this article is divided into five parts.In the first part,the first chapter of this article,the author systematically expounds the attributes(including ethical attributes and legal attributes)of intelligent robots.In the second part,the second chapter of this article,the author proposes the criminal risks and challenges we face in the era of artificial intelligence.In the third part,the third chapter of this article,the author analyzes the positions and ideas that criminal law and criminal law scholars should take in order to solve the criminal risks and challenges in the era of artificial intelligence.In the fourth part,the fourth chapter of this article,the author expounds the theoretical and practical countermeasures to meet the challenges of the artificial intelligence era.In the fifth part,the fifth and sixth chapters of this article,the author expounds the reflection and the idea of system reconstruction on the criminal theory system and the punishment theory system in the artificial intelligence era.The five parts are intertwined and deepened from the phenomenon to the essence,strictly following the basic logical train of thought to raise,analyze,and solve problems.The first part(Chapter 1)is the ethical and legal issues in the era of artificial intelligence.It studies and elaborates the ethical and legal attributes of intelligent robots.First,the ethical attributes of intelligent robots.In the context of artificial intelligence technology innovation and big data support,robots have become highly intelligent,and have been able to implement behaviors independently within the scope of procedures designed and programmed by humans,and their degree of autonomy is rapidly increasing.The difference between an intelligent robot and a natural person is that it does not have a human life.The difference between an intelligent robot and an animal and a general machine is that it can replace the function of the human brain to a certain extent.Therefore,from the legal attribute,the intelligent robot can be positioned as a “human worker”.It is foreseeable that the relationship between natural persons and intelligent robots will gradually fade in the future,and the dominant position of intelligent robots is gradually emerging.We should face this social ethical phenomenon and consider giving the intelligent robot appropriate constitutional qualifications and status.Second,the legal attributes of intelligent robots.Intelligent robots have the basis of legal personality,and treating them as legal subjects will not have a fundamental impact on the concept of “person” in law.The concept of “person” in law is not static.With the changes of the times and the transformation of ideas,the connotation and extension of the “person” in the law has undergone continuous evolution and change.Therefore,the use of intelligent robots as a legal “person” seems to be in line with the trend of the times.In fact,although our country’s laws currently do not stipulate the legal personality and rights and obligations of intelligent robots,countries or organizations around the world have already conducted research or legislation on this topic.It can be seen that it is not completely impossible to give legal personality and rights and obligations to intelligent robots in legislation.There are forefront legislative activities of the pioneers in the world.The second part(Chapter 2)is the criminal law problem and challenge that we face in the era of artificial intelligence.It studies and elaborates the criminal risk in the era of artificial intelligence.From the perspective of “longitudinal”,in the era of ordinary intelligent robots,the era of weak artificial intelligence and the era of strong artificial intelligence,there will be significant differences in the types and sizes of criminal risks faced by society;from the perspective of “horizontal”,in today’s weak artificial intelligence era,traditional crimes will undergo “quality change” and “quantity change.”First of all,from the perspective of “longitudinal”: artificial intelligence technology brings many criminal risks to human beings while bringing benefits to human beings,such as endangering national security and public safety,infringing on citizens’ personal and property rights,and undermining economic order and social order.However,it should be noted that there are essential differences in the criminal risks that different types of intelligent robots bring to human society.As far as ordinary robots are concerned,they are not different from general tools when they are used as criminal tools.When ordinary robots are used as criminal objects,they may affect the nature of crimes due to the characteristics of ordinary robots.As far as weak intelligent robots are concerned,they can only implement behaviors within the scope of procedures designed and programmed by humans.The behavior is essentially an extension of the developer or user behavior.The consequences of the behavior should also be attributed to the developer or user.For the criminal acts involved in most of the criminal risks,the Criminal Law and related judicial interpretations can be effectively regulated.However,we should also see that the spurt development of artificial intelligence technology and the lag of law have also formed a situation of disharmony,and the harm of “unable to follow” has emerged in some areas.The corresponding obligations should be set for the developers and users of weak intelligent robots,and the criminal responsibility path of the two should be clarified.As far as strong intelligent robots are concerned,when intelligent robots go beyond the design and programming procedures and implement behaviors under the control of autonomy and will,they are completely beyond the scope of tools.Because the strong intelligent robot at this time has independent consciousness and will,its behavior is no longer an extension of the behavior of the developer or user,and even fundamentally contrary to the purpose of the developer or user,the consequences of the behavior cannot be of course blamed on the developer or user.In this state,it is entirely possible for a strong intelligent robot to make independent decisions under the control of autonomy and will and to implement serious socially harmful behaviors beyond human control.Although the current criminal law has not yet been regulated,it should be regarded as the subject of criminal responsibility at the level of the law,and special penalties should be established.Secondly,from the perspective of “horizontal”,traditional crimes in the artificial intelligence era may will undergo “quality change” and “quantity change”.First,it may cause some of the harms of traditional crimes to “quantify”.On the one hand,from the “broadness” of crime hazard,artificial intelligence technology may make the coverage of criminal behavior more “wide”;on the other hand,from the “depth” of crime hazard,artificial intelligence technology can make products comprehensive.Artificial intelligent technology can also make criminal tools and criminal means more intelligent,so the social harm caused by criminal behavior may be more “deep”.Second,it may cause “constituent changes” in traditional crimes.On the one hand,the application of artificial intelligence technology may lead to new forms of crime;on the other hand,intelligent robots may be separated from human control for various reasons,and then independently implement criminal acts that seriously endanger society.The third part(Chapter 3)is the position and concept of criminal law in the era of artificial intelligence.It studies and elaborates on the legitimacy and appropriateness of criminal law in artificial intelligence crime and the criminal law concept that should be upheld in the artificial intelligence era.First,regarding the legitimacy and appropriateness of criminal regulations involving regulating artificial intelligence crimes.Faced with the risks that may be brought about by the development of artificial intelligence technology,the criminal law should intervene in time to protect the stability and security of the society.This is the legitimacy of criminal law involving regulating artificial intelligence crime.At the same time,the criminal law should not regard artificial intelligence technology as flood or a beast,prohibiting or hindering its development.Criminal law should not become a stumbling block in the development of artificial intelligence technology.This is the appropriate requirement for criminal law to deal with artificial intelligence crimes.Second,the concept of criminal law that should be upheld in the era of artificial intelligence.The existing criminal law stipulates that it is difficult to properly solve the emerging problems in the artificial intelligence era.In order to better play the protective function of criminal law,and promote the healthy development of artificial intelligence technology and the stability and prosperity of society,we need to establish a forwardlooking criminal law concept and establish and improve the criminal legislation and judicial system that meet the requirements of the era of artificial intelligence.The forward-looking concept of criminal law is different from the science fiction novel lacking a reliable scientific basis.It is not the same as blindly expanding the criminal circle,nor will it lead to the generalization of punishment.The forward-looking criminal law concept can reserve the necessary interpretation space and buffer space for the criminal law system involving regulatingartificial intelligence crimes,and avoid the revision of criminal law too frequently.The criminal law provisions can not only respond to current risks,but also adapt to future development,thereby enhancing and extending the vitality of criminal law provisions.The fourth part(Chapter 4)is a countermeasure to cope with the challenges of the artificial intelligence era.It studies the path of artificial intelligence criminal law.In view of the “vertical” and “horizontal” criminal risks faced by the society in the era of artificial intelligence,different artificial intelligence criminal laws and regulations are determined.First,from the perspective of “longitudinal”,different criminal regulations should be adopted for criminal risks at different stages of the artificial intelligence era.When identifying the criminal liability of an ordinary intelligent robot crime,it is necessary to clarify the difference between it and the traditional tools.That is,the human will be embodied in the ordinary intelligent robot through the program.The will embodied by the ordinary intelligent robot is the human will,so the ordinary intelligent robot can be the target of fraudulent crimes.For weak intelligent robots,on the one hand,it is necessary to establish corresponding risk prevention obligations for developers and users of weak intelligent robots.Ethical norms that have a major impact on society can be raised to legal norms.If relevant personnel violate the corresponding legal norms and cause serious harm to society,they can be investigated for criminal responsibility.On the other hand,weak intelligent robots are used as criminal tools.In the case of a weak intelligent robot committing a crime,if the developer and the user are not the same person,and there is no conspiracy between the developer and the user,it is necessary to subdivide the different situations to clarify the criminal responsibility sharing between the developer and the user.For strong intelligent robots,strong intelligent robots that implement behaviors outside the scope of design and programming are treated as behavioral subjects and members of society,and penalties are imposed on the behaviors that seriously endanger society.It is the regulation of strong artificial intelligence era.At the same time,the form and specific composition of the joint crime may also have different manifestations.The strong intelligent robot cannot form a joint crime with the developer,but there is a possibility that the strong intelligent robot and the user or the strong intelligent robot constitute a joint crime.Second,from the perspective of “horizontal”,different criminal regulations should be adopted for different types of crimes in the era of artificial intelligence.According to the current criminal law regulations on artificial intelligence crimes,we can classify artificial intelligence crimes into artificial intelligence crimes that can be regulated by the current criminal law,artificial intelligence crimes that are inadequately regulated by the current criminal law,and artificial intelligence crimes that cannot be regulated by the current criminal law.There are three types of artificial intelligence crimes.Firstly,in view of the artificial intelligence crimes that can be regulated by the current criminal law,there may be problems that are too vague,and we need to improve the relevant judicial interpretations so as to provide a comprehensive and accurate evaluation of such crimes.Secondly,in view of the artificial intelligence crimes that are inadequately regulated by the current criminal law,we need to adjust the constituent elements of the relevant crimes,and incorporate the updated behaviors into the scope of adjustment of the criminal law provisions.Thirdly,we need to establish new charges to regulate the new serious social harms in the current era of weak artificial intelligence.According to the different circumstances of the developer or the user’s intentional or negligent crimes,the crime of misuse of artificial intelligence products and accidents of artificial intelligence products were added.The fifth part(Chapter 5 and Chapter 6)is the reconstruction of the criminal law theory in the era of artificial intelligence,including the reflection of the criminal theory system and the reconstruction of the penalty theory system in the era of artificial intelligence.First,the thinking about the criminal theory system in the era of artificial intelligence.In view of the criminal theory problem in the era of artificial intelligence,the existing crime theory should be considered and reconstructed.First of all,it is about the subject of criminal liability of strong intelligent robots.Whether a strong intelligent robot with autonomy and will can become the subject of criminal responsibility,there are two opposite views in the academic world: affirmative and negative.Intelligent robots have been completely different from the swords of the cold weapon era and the guns of the hot weapon era.Even weak intelligent robots may get out of control in some cases.In this context,if we still insist that intelligent robots only have tool attributes,even when they are completely out of human control,they can only be treated as ordinary tools,which is obviously out of date and impractical.Since a strong intelligent robot may perform its behavior under the control of autonomy and will,we should reexamine the essence of a strong intelligent robot with criminal law theory.If it has all the necessary elements related to the subject of criminal responsibility,it should be criminalized.It should be noted that “intelligence” is only available to natural talents,and no other animal or item has “intelligence”.The difference between intelligent robots and general machines lies in the fact that they have the “intelligence” of natural talents,that is,artificial intelligence is actually that natural people create the "intelligence" that only human beings have.The "free will" based on "intelligence" determines the existence of independent identification and control capabilities.If the "intelligence" of intelligent robots fully meets or exceeds the intelligence of natural persons,its free will seems to be a self-evident conclusion.Strong intelligent robots have free will,and they have independent identification and control capabilities,and should be recognized as the subject of criminal responsibility.Secondly,it is about the identification problem of the subjective crime of the developer involved in artificial intelligence crime.Since weak intelligent robots and strong intelligent robots may play an autonomy within the scope of design and programming,and in the middle of the user’s use behavior,what should be the result of serious harm to society? This issue is worth exploring.The subjective crime of the developer involved in the artificial intelligence crime is determined.It includes two aspects: one is that the developer is guilty of artificial intelligence crimes;the other is that the developer has not any sin in the artificial intelligence crime.The main role of the former aspect is to prevent the improper punishment of the developers of intelligent robots and hinder the development of artificial intelligence technology;the main role of the latter aspect is to prevent developers from deliberately using intelligent robots as a tool to replace their criminal behavior or to cause serious harm due to negligence.Robots cause serious harm to society,thereby reducing the risks that artificial intelligence technology can pose to society.Accurately identifying the subjective crimes of developers involved in artificial intelligence crimes is conducive to the implementation of the basic principles of criminal law that are compatible with crimes,avoiding felony convictions or minor offences,and is conducive to promoting the development of artificial intelligence technology and preventing criminal risks involving artificial intelligence.When the developer designs an intelligent robot with the main purpose of committing criminal behavior,the developer’s subjective guilt should be deemed to be direct intentional for all the serious social harm caused by the intelligent robot.When the developer designs an intelligent robot with the main purpose of implementing non-criminal behavior,the result of serious harm to the society caused by the intelligent robot,when the developer violates the duty of care and has the criminal law clearly stated,the subjective crime should be recognized as criminal negligence,and should be based on the "intelligent" degree of intelligent robots to determine the identification criteria of the developer’s criminal negligence.When the above-mentioned hazard result is caused by ordinary intelligent robots,the type of negligence of the developer is direct negligence;when the above-mentioned hazard result is caused by the weak intelligent robot,the management negligence theory should be used to determine the processing standard of the developer’s criminal negligence;When a strong intelligent robot is created,the supervisory negligence theory should be used to determine the processing standard of the developer’s criminal negligence.In the era of weak artificial intelligence,the dynamic intelligence of weak intelligent robots is integrated into the behavior of natural people;in the era of strong artificial intelligence,strong intelligent robots implement behaviors autonomously and achieve complete control over the behavior process.The impact of "behavior" in the new scene of artificial intelligence on the behavioral connotation of criminal law is mainly reflected in whether the behavior of intelligent robots conforms to the first condition of behavior in criminal law(that is,whether it is subject to human consciousness).If the above-mentioned behaviors in the new artificial intelligence scene are included in the scope of behavior in the criminal law,there will be two dilemmas.First,whether the active integration of weak intelligent robots in the behavior process will affect the evaluation of behavioral power by human consciousness.The first condition for determining an act as a crime in the criminal law is “to be dominated by human consciousness”.The usual understanding of “dominance” is that human consciousness plays a 100% influence and role in the behavior process.However,when the dynamic intelligence of weak intelligence robots is integrated into the behavior process,the influence and effect of natural people on behavior seems to be no longer 100%.Second,whether the first condition of the behavior in the criminal law "to be dominated by human consciousness" must limit the subject to "person",can it be other subjects with similar consciousness to people? If the behavior of a intelligent robot beyond the scope of human design and programming is included in the scope of criminal law behavior,it means that the intelligent robot has the same qualification as the subject of the behavior in the criminal law.This is even more an impact on the connotation of behavior in criminal law in traditional criminal law theory.It should be noted that the behavior of incorporating the dynamism of weak intelligent robots is still carried out under the control of human consciousness and will;the behaviors of selfimplemented by strong intelligent robots and the behaviors carried out by natural persons under the control of autonomy and will are the free will of the actors.Taking the legal basis of the behavioral connotation in criminal law as the basis for judging,the above two behaviors should be included in the scope of behavior in criminal law.This is the due dilation of the meaning of behavior in criminal law in the era of artificial intelligence.Finally,the problem of attribution of artificial intelligence crimes.The solution of attribution problems in artificial intelligence crimes depends on the identification of the causal relationship between the artificial intelligence crimes and the serious harm to the social results.The solution to the problem of artificial intelligence crime depends on the behavior.On the basis of following the general law of causality determination in criminal law,the condition of dual screening should be combined with the characteristics of artificial intelligence crime as the attribution criterion of artificial intelligence crime.The condition of dual screening is based on the conditional theory,which draws on the rational component of objective imputation theory and establishes the criterion of causality.In the case of artificial intelligence crimes,the principle of strict liability should not be adopted,and the subjective crime of the perpetrator should be taken as an essential requirement for the criminal responsibility of the perpetrator.Second,on the reconstruction of the penalty system in the era of artificial intelligence.On the basis of determining that the strong intelligent robot has the qualification of criminal liability subject,when the intelligent robot implements the behavior that seriously endangers the society outside the scope of the design and preparation procedures,it should be punished.The existing penalty system in China consists of life punishment,free punishment,property punishment and power punishment.The object of punishment cannot punish strong intelligent robots.Reconstructing China’s penalty system and incorporating strong intelligent robots into penalties is in line with the purpose of penalties,and it also meets the needs of the development of artificial intelligence era and does not violate basic jurisprudence.It is recommended to add penalties for de-duplication data,modification procedures,permanent destruction,etc.that can be applied to strong intelligent robots,and add penalties for property punishment or power punishment for strong intelligent robots when conditions are ripe.The inclusion of strong intelligent robots in the scope of penalties is essentially the recognition of the social membership of a strong intelligent robot,which is determined by the degree of participation in human production and life and the economic and social status of strong intelligent robots. |
PDF Full Text Request