Research On The Legal Regulation Of Algorithm

In a broad sense,Algorithm is a problem-solving program,but due to the continuous development of data and computing power,the evolution of basic mathematical theory,breakthroughs have been made as the essence of artificial intelligence.Multi-type algorithms based on machine learning are in the news It is used in many scenarios such as push,search engine,e-commerce,credit score,and judicial judgment.The improvement of algorithm performance and the popularization of applications have improved the operating efficiency of society,but they may cause many risks due to internal opacity,limited autonomy,and difficulty in accountability.These characteristics distinguish algorithms from ordinary technologies,and also bring us regulatory difficulties.When self-regulation,market regulation,and ethical regulation cannot effectively deal with the risks and impacts of algorithms,legal regulation has become the most direct and effective Regulatory means.It is necessary to take the iterative update of algorithm technology as the background,make the legal regulations of algorithms as the main research object,highlight the limited autonomy of current algorithms,and construct a legal regulatory system that is in line with the current status of algorithm applications in China.With the development of the Internet,online activities have become more and more important in people's daily lives.The algorithm will provide us with more rapid and efficient push results,and it will also affect our horizons,judgments and even final choices.It can be said that the network has formed a path dependence on the spontaneous and spontaneous expansion order formed in its development.Obviously,the concept of technological neutrality and technical innocence that algorithm users uphold has formed a strong inertia.The consequences of this inertia inevitably conflict with the rights of citizens and the orderly development of society.During this process,risks such as algorithm discrimination,algorithm domination,and even social trust dissolution will be formed,and individuals will gradually evolve into a summary of data before the algorithm,which will inevitably be disciplined by the algorithm,the difficulty of privacy protection will gradually increase,and the learning characteristics of the algorithm It will also lead to an increased digital divide.These social risks,as well as the adverse effects on individuals,require a strong response from the law.In terms of the theoretical basis of the legal regulation algorithm,it is necessary to make clear that algorithms can only become the object of legal regulation and not the law itself.We should recognize the legal attributes of algorithms and position them as artificial objects with limited autonomy.At present,the European Union and the United States have formed algorithmic regulation models centered on data protection and algorithmic responsibility,but both have certain limitations.At this stage,China's regulatory model for algorithms is still responsive,focusing on the regulation of results and ignoring process control.The specific manifestations are that the personal data protection legislation is relatively decentralized,the platform liability regulations are not clear enough,and the legal system of regulatory algorithms is insufficient.Therefore,China needs to build a systematic legal regulation system.Furthermore,it adopts legal governance and natural technological governance,positioning risk prevention as the goal to be achieved by system design,and using science and technology ethics as its internal guidance.The system construction of the legal regulation system covers the entire process of algorithm application,mainly including review and supervision before application of the algorithm,data protection during application,and legal accountability when consequences occur after application.First of all,in the study of algorithm review and supervision of specific regulatory measures,the algorithm's hierarchical classification system needs to be improved.The main application scenarios of current algorithms are concentrated in two areas,one is the business area and the other is the public decision area.After the risk level of the algorithm is determined,the corresponding transparency boundaries are determined according to different levels.Completely opaque algorithms will form a black box,while fully open algorithms cannot protect the enthusiasm for innovation of algorithm developers.Therefore,establishing a sound supervision mechanism by clarifying the transparency of the algorithm can effectively prevent problems from occurring at the stage before the algorithm is applied.Secondly,through the research on the data rights system,strengthen the legal guarantee in the application process of the algorithm.Clarifying the definition of data ownership is critical for limiting algorithms,and algorithms without data will not work.However,Chinese law does not clearly define the increasingly diverse data forms,and there is a contradiction between data controllers and data subjects in the collection and use of data.Therefore,while emphasizing the use of data by Internet companies,the protection of personal data rights cannot be ignored.Giving data subjects the right to portability of the data and strengthening the protection of the right to be forgotten can effectively deal with data out of control brought about by the application of algorithms.Finally,the legal consequences of the application of the algorithm and the question of how to hold it accountable are studied.Strengthen the interpretability of the algorithm and protect the data interpretation power of the data subject.It can review the process of algorithm operation,find out the specific causes of legal consequences,and reasonably allocate responsibilities to complete legal accountability.The legal regulation methods designed for the entire process of algorithm application can only solve the problems presented in the short and medium term,and the proposal of digital human rights can affect the algorithm from a more macro perspective,thereby achieving the purpose of regulating the algorithm at the source and implementing the algorithm.The long-term goal of regulation.Digital human rights require people-oriented in the process of digital technology development.On the one hand,the value advocated by the digital human rights protection mechanism is the value that scientific and technological development should follow.On the other hand,the coverage of digital human rights is extremely wide,and in the face of a wide range of applications of digital technology represented by algorithms,digital human rights can be constrained.Protecting digital human rights requires strengthening the human rights concepts of algorithm developers,strengthening the moral acquisition capabilities of algorithms,enhancing the participation of digitally disadvantaged groups,and fostering digital literacy and algorithm awareness in society as a whole.The proposal of digital human rights can lead the construction of legal regulatory system and the concept of algorithm design.Furthermore,combined with China's ongoing community of human destiny,digital human rights can enhance China's right to speak in the fields of science and technology legislation and data governance.