人工智能时代下的算法歧视及其治理研究

Algorithmic Discrimination and Its Governance in Era of Artificial Intelligence

随着大数据与人工智能技术的飞速发展,作为人工智能的核心技术之一,算法被广泛应用于人类社会中,其在推动产业经济发展与便利人类生活的同时,亦带来严重的算法歧视问题。算法歧视不仅在直接层面上侵害特定主体的合法权益、经营者的竞争性利益、民众的基本人权,也在间接层面上加深既有偏见,加剧数字鸿沟,不利于人工智能产业的长远发展,故亟需对其进行规制。通过对算法歧视肇因追本溯源可知,算法歧视乃算法开发者内隐偏见嵌于算法、数据偏差、算法黑箱等信息不对称的结果。算法歧视作为技术领域的法律问题,在借鉴域外模式的同时需要采取法律、技术、社会规范等多元规制手段,多管齐下,以达到对算法歧视治理的同频共振的效果。

With the rapid development of big data and artificial intelligence technology,algorithms,as one of the core technologies of artificial intelligence are widely used in human society.While promoting the development of the industrial economy and facilitating human life,it also brings serious problems of algorithmic discrimination.Algorithmic discrimination not only infringes the legitimate rights and interests of specific entities,operators’competitive interests,and people’s basic human rights at the direct level,but also deepens existing prejudice and aggravates the digital gap,which are not conducive to the long-term development of the artificial intelligence industry and need to be regulated urgently.By tracing the causes of algorithmic discrimination,it can be seen that algorithmic discrimination is the result of algorithm developers’implicit bias embedded in algorithms,data deviation and algorithm black boxes.Algorithmic discrimination is a legal issue in the technical field.Based on referring to the extraterritorial mode,the regulation of algorithmic discrimination needs to adopt multiple means such as law,technology and social norms to achieve the coordinated effect of algorithmic discrimination governance.