预测算法与社会公平——以“机会信息”访问为视角

Predictive Algorithms and Social Equity:A Perspective from Access to“Opportunity Information”

[目的/意义]通过研究和探讨对个人权益有重大影响的“机会信息”(如就业机会、教育机会等)的访问,为数据主体基本权利的法律保护提供建议。[研究设计/方法]以脸书(Facebook)“纳米定位”广告为例,结合现如今网上招聘的运作模式,阐明预测算法如何控制对以就业信息为代表的机会信息的访问并可能加剧甚至恶化现有的不平等模式,进而损害社会的公平公正。[结论/发现]当互联网平台服务商设计定位算法来确定哪些广告投放给哪些用户时,它就已经超越了网络服务提供者的角色,变成了网络内容提供者,应该对此影响承担相应责任。[创新/价值]提出并定义了“机会信息”的概念,为阐明算法的公平性问题提供了新角度。

[Purpose/Significance]In order to provide advice on the legal protection of the fundamental rights of data subjects,this study analyzes and discusses access to"opportunity information"(such as employment opportunities,educational opportunities,etc.)that significantly impact individual rights.[Design/Methodology]Taking Facebook’s"nano-targeting"advertisements as an example,combined with the operation mode of online recruitment,this paper explains how predictive algorithms control access to opportunity information,especially employment information,which may exacerbate or even worsen existing patterns of inequality,and undermine social fairness and justice.[Findings/Conclusion]The results show that when an Internet platform service provider designs a positioning algorithm to determine which advertisements are served to whom,the provider has surpassed its role as a network service provider and has become a network content provider,and should be held responsible for this impact.[Originality/Value]The concept of"opportunity information"is proposed and defined in this paper,which provides a new perspective for clarifying the fairness of algorithms.