上海大学学报(社会科学版) ›› 2025, Vol. 42 ›› Issue (6): 17-28.

• • 上一篇    下一篇

端侧人工智能时代的人脸识别技术治理转型

  

  1. 中国人民大学 法学院
  • 出版日期:2025-11-11 发布日期:2025-11-27
  • 作者简介:
    黄尹旭,中国人民大学法学院副教授,法律科技与社会治理实验室研究员,国家治理大数据和人工智能创新平台研究员,法学博士。研究方向:数字经济、数字治理、数字法学。
  • 基金资助:
    国家社会科学基金重点项目(24AZD021)

Governance Transformation of Facial Recognition Technology in the Era of On-Device Artificial Intelligence

  1. Law School, Renmin University of China
  • Online:2025-11-11 Published:2025-11-27

摘要:

人脸识别技术因应用场景异化面临多元风险,衍生出治理成本与收益失衡的结构性矛盾,其安全验证功能的单向度发展超越社会价值理性的边界。人脸识别技术的核心价值并非局限于安全验证场景,其在构建自然世界与数字空间的身份连接纽带方面具有独特优势,有助于实现物理身份与数字身份的同一性认定,并保留人类生物特征的自然属性。依循“风险—收益”匹配场景引导性治理思路,可将人脸识别技术的场景功能定位从安全工具转向数字基础设施,进而探索构建以必要性原则和场景适配性为核心的动态治理机制,重塑技术应用与社会发展的关系,构建以人为本的伦理规范约束。

关键词:

Abstract:

Facial recognition technology (FRT) faces diverse risks arising from the alienation of application scenarios, giving rise to a structural contradiction between governance costs and benefits. Its unidirectional development, primarily as a tool for security verification, has extended beyond the bounds of social value rationality. The core value of FRT is not confined to security verification; rather, it possesses a unique advantage in forging identity linkages between the natural world and digital space, enabling consistent recognition of physical and digital identities while preserving the natural attributes of human biometric features. Following a scenario-guided governance approach that aligns risks and benefits with specific use-case scenarios, this study proposes a strategic shift in the functional positioning of FRT—from a simple security tool to a form of digital infrastructure. This transition would allow for the exploration of a dynamic governance mechanism centered on the principles of necessity and scenario-appropriateness, thereby reshaping the relationship between technological applications and social development, and constructing human-centered ethical norms.

Key words:

中图分类号: