人脸识别技术不合理扩散的伦理风险与法律规制--兼论动态同意规制模式的建构  

Ethical Risks of Unreasonable Proliferation of Face Recognition Technology and Legal Regulation--Also on the Construction of Dynamic Consent Regulation Model

作　　者：汤建华[1] Tang Jianhua

机构地区：[1]华中农业大学文法学院,湖北武汉430070

出　　处：《新疆社会科学》2022年第3期135-144,180,共11页Social Sciences in Xinjiang

摘　　要：人脸识别技术在社会生活中的应用越来越广。人脸识别技术早期应用目的是维护社会安全,其不合理扩散引发了伦理风险,数据主体的隐私泄露侵犯人的自由和主体性,算法歧视危及社会公平正义。对于人脸识别技术的扩散需要进行法律规制,针对隐私保护,需要在法律上构建知情同意原则、最小必要原则和不伤害原则;针对算法歧视需要在法律上构建算法中立原则和公正原则。其中,知情同意原则具有优先性,围绕知情同意原则的具体建构,欧盟立法采取了“原则—例外”的特别同意模式,即原则上禁止处理人脸数据,除非数据主体同意以及法律允许。我国采取了“隐私权+个人信息”双重保护模式,体现了宽松同意模式。两种模式均无法逃脱价值冲突的困境,需要建构一种具备可操作性的动态同意模式。Face recognition technology is being used more and more widely in social life.The ear-ly application of face recognition technology aims at social security,but its unreasonable prolifera-tion raises ethical risks,the privacy leakage of data subjects violates human freedom and subjectivi-ty,and algorithmic discrimination endangers social justice.The proliferation of face recognition tech-nology needs to be legally regulated.For privacy protection,the principles of informed consent,min-imum necessary and do no harm need to be legally constructed,and for algorithmic discrimination,the principles of algorithmic neutrality and fairness need to be legally constructed.Among them,the principle of informed consent has priority,and the EU legislation adopts a special consent model of"principle-exception"around the specific construction of the principle of informed consent,i.e.the processing of face data is prohibited in principle,unless the data subject agrees and the law allows it.China has adopted a dual protection model of"privacy+personal information",reflecting a leni-ent consent model.Both models cannot escape the dilemma of conflicting values and require the con-struction of a dynamic consent model with operability.

关 键 词：人脸识别技术 伦理风险 知情同意原则 动态同意模式 

分 类 号：D922.16[政治法律—宪法学与行政法学]