The Institutional Logic Of Legal Allocation Of Risk For Artificial Intelligence Technolog Applications

Since the inception of artificial intelligence(AI)technology,human society has e ntered a new phase of development.However,it is important to note that today’s AI te chnology is still considered a weak AI,and there remains a significant gap between it and strong AI,which can learn and think independently.In this context,a careful analy sis reveals that the logic of AI technology is a reflection of the "micro" power operatio n process,guided by the concept of operationalism and dependent on traditional public power.This not only makes AI technology challenging to control,but also gives it the ability to regulate human beings in turn,posing challenges to social relations,social o rder,and the value objectives of social institutions.The challenges brought by the application of AI technology have given rise to the risk of AI technology application,which is an artificial risk that can affect all subjects.These risks include big data risks,risks related to AIGC technology,and automated de cision-making risks.These risks have unique characteristics that distinguish them from ordinary artificial risks,such as the non-fully human nature of automated decision-ma king,the particular uncertainty of risk occurrence,and the variability that exposes diff erent subjects to different risk problems.The use of AI technology can either empower or deactivate public power,leading to the refinement and privatization of the power o f the state.It not only has the ability to gradually erode individual subjectivity but also violates many individual rights.Presently,there is a set of risk allocation systems in China that deal with the risk of AI technology application,considering the impact on power and rights.These syste ms include the system of power allocation without authorization,the system of legal b ehavior of meaningful autonomy,and the system of accountability of the "behavior-lia bility" mechanism.However,these systems face a risk allocation dilemma.In public l aw,the system of allocation of power without authorization faces the dilemma of desir ability in terms of effectiveness,economy,and cheapness,making it challenging to set power grant,boundary,and disclosure.In private law,the system of meaningful auton omy of legal acts faces a dilemma in terms of freedom and equality of fundamental rig hts due to information asymmetry,making it difficult for individuals in the risk of AI t echnology application to understand meaningful consent and the burden of the right of meaningful "autonomy." The common dilemma between public law and private law is that the recourse system of the "act-liability" mechanism requires the subjectivity of t he responsible subject,making it difficult to operate properly,leading to difficulty in p roperly operating the "conduct-responsibility" system.To overcome the above dilemma,new methods are necessary to achieve fair distr ibution of risks.In public law,the principles of publicness,checks and balances of po wer,due process,and modesty and restraint can be used as guidance to establish a pow er allocation list system,effective power responsibility list system,economic power co ordination and supervision system that encourage participation of multiple subjects,an d a system to monitor the operation of public power.Additionally,an inexpensive publ ic procedure for monitoring the operation of state power and a unidirectional system f or assigning responsibility using a prohibition-based approach in conjunction with a ri ghts-based approach is necessary.In private law,the principles of freedom,equality,anthropocentrism,and the supe riority of the weak can be used as guidance to establish a free and equal "sandbox" env ironment that narrows the information gap between subjects in risk allocation before ri sk allocation.Establishing a fair and reasonable risk allocation system before risk ente rs the traditional legal system,a legal protection system to examine whether AI-related interests can become a right and be protected by the anthropocentric principle,and a l iability system that can dynamically allocate risks to protect vulnerable groups in acco rdance with the principle of the priority of the weak are also crucial.