On The Positive-sum Model Of Legal Liability Allocation By Intelligent Algorithms

The goal of the algorithm’s legal liability allocation should be "positive sum",that is,to build trust,cooperation and win-win man-machine relationship.There are three traditional distribution modes of algorithmic legal liability: strict liability mode,indirect liability mode and platform liability mode,which contain efficiency principle and egoism.The problems of universality and reasonable expectation exist in the traditional algorithm’s legal liability distribution,which show that the traditional algorithm’s legal liability distribution model does not meet the requirements of the positive-sum objective.Is there a distributive model that achieves "positive-sum" goals in the face of traditional model dilemmas?This paper proposes a "positive-sum pattern" in response to the requirement of positive-sum goals.Firstly,the positive-sum model contains the principle of fairness and altruism,which realizes the positive-sum goal from the theoretical level and finally achieves mutual benefit and win-win situation.Secondly,through Coase theorem,Posner theorem and incentive mechanism to achieve positive-sum goal in practice.Finally,it discusses the necessity of positive-sum model through social contract.Defend the feasibility of positive-sum model through utilitarianism.Finally,the legal construction of positive-sum model.The realization of positive-sum model from the theoretical dimension to the practical dimension is centered on the three positive-sum goals of "trust,cooperation,mutual benefit and win-win".Firstly,appropriate transparency principle and individual’s right to know should be taken as legal principles to embed the cooperation between the two parties.Secondly,self-imputation and external imputation,scene imputation and refined imputation are used as legal imputation to embed mutual trust.Finally,the algorithm qualified incentive system and disclaimer trigger mechanism as legal mechanisms to achieve mutual benefit and win-win.