Study On Criminal Imputation Of Crimes Involving Artificial Intelligence

With the continuous development and wide application of artificial intelligence technology(AI),many risks associated have gradually emerged in society.In the field of legal studies,especially criminal law,the criminal risk brought by AI has attracted widespread attention from scholars.Suddenly,AI criminal law research emerged and the AI criminal responsibility theory has become the most significant knowledge growth point.At the same time,some scholars question the research value of the topic“AI criminal risk and criminal law response”.It is believed by some scholars that the current research violates the common sense of human intelligence and is more an“academic show” to track social focus.This led to discussions and controversies about the criminal imputation of artificial intelligence crimes.The two parties debated whether the concept of artificial intelligence can be defined,whether artificial intelligence criminal risks exist,and whether artificial intelligence criminal law research can meet social needs.In fact,the analysis of the subject's ontological concept,attribute and value also constitute the prerequisite thinking for the specific study of the criminal responsibility of artificial intelligence crime.Only by fully responding to relevant queries,can the research on artificial intelligence criminal law,including the discussion of the criminal imputation of artificial intelligence crimes,be carried out.In this regard,the author believes: First,although attempts to unify the concept of AI in a scientific sense seem to be impossible,this does not prevent the description of AI to meets the needs of current legal research.Second,crimes involving AI have occurred,are developing,and will spread further.Finally,risk theory and AI criminal law research have many points of convergence,and the risk of the current society have been widely recognized in different disciplines.It is of social value to discuss AI from the perspective of risk society.Therefore,the artificial intelligence criminal law research focusing on the criminal imputation of artificial intelligence crimes has its foundation and value.The technical risk derived from AI at the fact level will first cause the imputation difficulty at the normative level.When the traditional criminal law cannot comprehensively regulate AI crimes,part of the criminal liability cannot be matched with certain subject.However,the difficulty of imputation does not exist in all types of crimes.In intentional crimes involving AI,the existing “Theory of Mistake” and“Theory of Principle” can completely avoid the generation of imputation gaps.But in AI-related negligence crimes,things are totally different.First,AI technology characteristics and technical risks contradict the imputation elements in negligent crime theory,which manifests as AI technology leads to impossibility of foresight and impacts the imputation fundation.Second,AI technology make the causality in crimes much more complicated and hinder criminal responsibility imputation.Third,the lack of law and regulation on AI technology makes it difficult to define the duty of care.Fourth,AI technology breaks the precedent that only individual shall be liable for a single act.Fifth,AI crime cannot be handled by traditional experience,which cause the difficulty in imputation.As a result,it has caused the dilemma of AI negligence crime.In this regard,scholars have proposed three theoretical approaches.One is the Approach of Interpretation,advocating that to abstract the foreseeability and set wide duty of care by the criminal law interpretation in order to avoid inability to foresee and avoid responsibility.The second is the Approach of Legislative,which advocates establishing the principle of strict liability in the general regulations through legislation,and adding abstract dangerous crime in the sub-rules to omit the investigation of causality,so as to smoothly find the subject liable.The third is the Approach of Countermeasures,which advocates giving AI the subjective status of criminal responsibility and setting up a penal system for it and assigning criminal responsibility between artificial intelligence and natural persons through the establishment of “AI Criminal Law”.However,all approaches will cause risks to the criminal law system(secondary risks).The abstraction of foreseeability,the marginalization of criminal damage,the omission of causality,and the localization of fault liability have impacted the integrity of the criminal law system,weakened the fairness of criminal liability,and undermined the legitimacy of the penalty system.Based on the specific analysis of the imputation dilemma in AI crime,the effective theoretical approach to imputation dilemma should reflect the following three basic positions.First,exclude AI from being the subject of criminal responsibility.Second,the rational attitude shall be abided by criminal law and its risk concept.Third,to meet the realistic needs of dealing with imputation dilemma.The theoretical transformation of the negligence criminal liability structure,which means taking normative and objective imputation mode rather than voluntary and subjective mode,is an effective and beneficial attempt in the process of choosing AI imputation dilemma solution.“If the theoretical core is free will of individual in the Imputation of Will,then the core consideration is the violation of norms by behavior and results in the Normative Imputation.” The Normative Imputation model takes the judgment of the risk jurisdiction as the logical premise and takes the individual's behavior to create and realize the risk not being allowed as the substantive content of the judgment of criminal responsibility,which is different from the Imputation of Will.The meaning of risk jurisdiction is that “It is necessary to preliminarily examine who has the jurisdiction to such unbearable risk when one use Normative Imputation to make a judgement in criminal negligence.” If the risk belongs to the individual's circle of competence,it is necessary to further determine whether the individual's behavior creates an unbearable risk,that is,whether the behavior creates a risk,and whether the behavior has a cause for exclusion.When the affirmative behavior has the possibility of disapproval,it is still necessary to examine whether the results caused by the behavior realize the unbearable risks,which means the risk is realized within the scope of the normative protect purpose,the risk creation has connection with the the realization of risks and the creation of risks at least significantly increases the possibility of realization of risks.In that way,the liability can be finally confirmed.There are several significant meanings of Objective Imputation Theory in the imputation model transformation and AI criminal negligence imputation gap filling.First,the objective imputation theory realized the transformation of imputation foundation from empirical facts to the norm so that the imputation judgment may get rid of the foreseeability and life experience in the risk society,which is difficult to definite.Secondly,the Objective Imputation Theory has realized the change of the Imputation from the Causation Theory to the Jurisdiction Theory,gave the causation relationship normative meaning and normative test model,and solved the imputation dilemma caused by the vague of causation in actual scenarios.Third,the Objective Imputation Theory realizes the transformation of imputation from vertical to horizontal dimension,broadens the idea of imputation,and makes the distribution of criminal liability between subjects in negligent crime a content that needs to be considered.Fourth,the Objective Imputation Theory use behaviours that conform to obligations to replace behaviours that violate obligations in order to examine the relationship between behaviours and results,which can make the imputation get rid of the causality.At the same time,expressing the existence of relationships with probability can also overcome the difficulty of imputation.At last,the Objective Imputation Theory can reasonably limit the improper expansion of negligent crime by examining whether the risk created by individual shall be protected by foreseeability clause,which can prevent the shrinking of AI technology while achieving effective control of AI crimes.