On The Legal Liability For Harm Caused By Artificial Intelligence

When artificial intelligence and legal responsibility are put together,this strange combination will undoubtedly give people the first impression of science fiction and even magic.But in the final analysis,impression is only a vague sign of objective things mirroring in the mind,and impression is far from the essence of things.In recent years,artificial intelligence technology has already penetrated into our daily lives,has changed and will eventually reform our way of life.The law should be the protagonist of this change,timely regulate the corresponding risks and continue to maintain its control in guiding,planning and controlling our lives.The essential difference between artificial intelligence and other technologies is that we entrust artificial intelligence to make judgments on their behalf,thereby freeing humans from the burden of judgment and choice.This means that artificial intelligence will have certain judgment capabilities.In the weak artificial intelligence stage,this ability is a "predetermined judgment" made by a pre-coded program,and in the strong artificial intelligence stage,this ability is close to the "autonomous judgment" of artificial intelligence.Therefore,when artificial intelligence causes damage,how the legal mechanism will deal with this part of "out of control" behavior has become a difficult and important issue.There are two ways to solve this issue:attribute the damage to the artificial intelligence that caused the damage or attribute the damage to other individuals.The prerequisite for the application of the former is that the artificial intelligence has the qualification to be held accountable,and the prerequisite for the application of the latter is that the damage should be borne by the individual.If it is believed that it is desirable to attribute the damage to artificial intelligence,it is necessary to discuss whether artificial intelligence can be the subject of responsibility,that is,to discuss the subject qualification of artificial intelligence.The jurisprudence of subject qualifications is based on the rational ability of human beings.The key to distinguishing the "person" of the subject from the "thing" of the object is that the subject has the ability to act and choose freely.And free will is also the basis of responsibility.The above theory tries to show that if artificial intelligence possesses free will and rational ability,it is sufficient to include it as a legal subject to make it independently responsible,but to attribute the harmful results caused by its"self-judgment" "behavior" to others The individual is not reasonable.If it is believed that it is desirable to attribute the damage to other subjects,it is necessary to demonstrate that the source of the harm caused by artificial intelligence comes from other subjects,that is,other subjects still make their own judgments and choices instead of artificial intelligence.Due to the different nature of artificial intelligence at different stages,the methods of imputation should also be discussed separately.The process of weak artificial intelligence making judgments is just a fixed result of a pre-loaded program after regular screening of a huge database.Similar to presupposing n kinds of solution paths for n kinds of situations,the artificial intelligence at this stage is completely"according to the rules" and will not derive the n+1th solution path beyond the original program setting,and therefore it does not Similar to human reason and the ability to choose independently,its judgment and actions are fundamentally controlled by designers,producers or users.In this case,it is reasonable to blame other individuals for the damage caused by artificial intelligence.However,strong artificial intelligence or super artificial intelligence that has surpassed the weak artificial intelligence stage has rational ability and free will,and its judgments are almost no different from human "autonomous judgments",and other subjects have no control over their judgments.Obviously,the responsibility for the damage should be attributed to the artificial intelligence itself.Based on the above argumentation,it can be seen that the responsibility for damage caused by weak artificial intelligence should be attributed to other responsible entities,and the responsibility for damage caused by strong artificial intelligence should be attributed to the strong artificial intelligence entity itself,and the legal liability system for artificial intelligence should be constructed based on this idea.