| Algorithm is the core of artificial intelligence.Algorithm technology not only brings development opportunities,but also brings new challenges to the construction of legal rules and order in today’s society,exposing many risks including algorithm discrimination.Algorithmic discrimination is a further extension of traditional discrimination in the new era.It has the characteristics of concealment,systematic and universality different from traditional discrimination.It is manifested in the fields of price,employment,race,gender and so on.Since algorithmic discrimination may lead to personal risks such as weakening of individual subjectivity,information cocoon room and damage to personal dignity,resulting in social risks such as uneven distribution of social resources and destruction of social fairness and justice,it is extremely necessary to regulate algorithmic discrimination by law.In the era of artificial intelligence,algorithm discrimination has gradually penetrated into all aspects of social life.Based on the analysis of legal regulation of price discrimination,it can be found that the current legal regulation of algorithm discrimination is also faced with practical difficulties,such as insufficient transparency of algorithm,difficulty in proving and safeguarding rights,unclear subject of supervision,lack of industry self-discipline and so on.The main reasons for these difficulties are the defects of the legal system,the application of algorithm scenarios and the public dependence caused by them,and the limitations of algorithm technology.For algorithmic discrimination,foreign countries have rich experience.At present,European countries have formed an algorithm regulation model centered on data protection,while the United States has chosen a legal regulation model centered on algorithm responsibility.Paying attention to the specific system design,adhering to the equal emphasis on risk prevention and post accountability,and promoting the transformation from extensive supervision to diversified supervision are important experiences that can be used for reference in the regulation of algorithmic discrimination abroad.In the era of artificial intelligence,the legal regulation of algorithm discrimination can consider three dimensions: before,during and after.In the ex ante dimension,we should establish the regulation principle oriented by risk prevention,pay attention to the governance of algorithm technology,and actively explore the mechanism of algorithm review and evaluation.At the event level,we should strengthen the situational standardization and supervision of algorithm application by law,rationalize restrictions on algorithm discrimination,and improve the regulatory role of regulators in algorithm discrimination.At the post event level,we should further improve the legal relief measures of algorithmic discrimination,which mainly includes the construction of a special public interest litigation mechanism and coping with it through autonomous industry supervision.The comprehensive application of these regulatory measures will help to prevent the adverse effects of algorithm discrimination,realize the effective guidance of algorithm development,and make the artificial intelligence supported by algorithm better benefit mankind. |
