论算法与法律行为的关系:制度影响与法律回应
CSTR:
作者:
作者单位:

同济大学 法学院,上海 200092

作者简介:

通讯作者:

中图分类号:

G90-052

基金项目:

国家社会科学基金一般项目“数字时代中算法消费者的公私协同保护体系研究”(24BFX076)


On the relationship between algorithms and legal acts: Institutional influence and legal response
Author:
Affiliation:

School of Law, Tongji University, Shanghai200092, P R China

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    大数据和人工智能的发展深刻地影响着人类的社会生活,改变了我们的生产和生活方式,也给法律规制带来了新的问题。我们从互联网社会走向算法社会,算法正在通过不同形式,不同程度地介入合同、侵权行为、公司治理、保险合同、证券交易等整个私法领域。在人工智能场景的算法社会中,形成了算法消费者群体。讨论算法如何影响法律行为效力的前提在于算法是否会影响行为人的意思表示。算法是针对数据的一系列程序逻辑,该程序逻辑是在自然语言基础上建构起来的。尽管算法具有一定的学习能力,可以通过大数据不断地进行深度学习,但这并不是认知层面的学习,不完全等同于人类学习。算法的本相是一种基于布尔代数的程序性“弹珠迷宫游戏”,其不符合民法中意思表示结构。通过自动化决策的算法系统作出的意思表示,仍是人类意思表示的延伸。算法使用者基于大数据和算法鸿沟形成了算法权力,对意思表示制度造成影响。算法不会对法律行为效力制度造成根本性挑战,但在错误、欺诈制度的具体构成要件中需要进行重新解释。在解释时,需要区分算法的使用人是表意方还是相对人的不同场景。在算法社会中,消费者向算法消费者演进,并作为一种消费者的新类型。为回应现有私法框架对于算法消费者保护不足的问题,应当调整《保险法》中如实告知义务的模式,《消费者权益保护法》需重塑经营者与消费者之间的权利义务;在商事金融领域,利用算法的金融机构将被赋予更高的注意义务。算法消费者保护应当考虑多维度的因素,一方面从个人信息保护体系和将信息信义义务作为企业社会责任等方面进行;另一方面构建算法透明度制度、算法备案制、算法伦理等规制算法的全方位监督体系。通过法律规制算法的实质是监管算法使用人(如算法型企业或平台)在网络交易或经营过程中如何规范与合理地使用算法。当前关于人工智能的研究,应回归学术理性,避免玄幻主义似的盲目跟风式。算法社会中算法消费者保护的核心与关键在于如何在个人信息保护、规制算法和算法企业发展之间寻求利益平衡点。

    Abstract:

    The development of big data and artificial intelligence has profoundly influenced the social life of human beings, changed our production and life style and led to a number of challenges for the law. We are moving from the internet society to the algorithmic society, with algorithms intervening in the overall private law field including contracts, torts, corporate governance, insurance contracts, and securities transactions in different forms and latitudes. In the algorithmic society of artificial intelligence, the group of algorithmic consumers formed. The premise for discussing how algorithms affect the effectiveness of legal acts is whether they affect person's intention manifestation. Algorithms are series of digital program logic constructed on the basis of natural language. Algorithms have a certain degree of learning ability, which can learn continuously and deeply through big data. However, it is neither learning in cognitive level nor human learning. The essence of algorithms is a kind of procedural Marble Maze Game based on Boolean algebra, which does not conform to the structure of intention manifestation in civil law. The intention manifestation made by the algorithmic system of automated decision-making is still an extension of human intention manifestation. Algorithm users form algorithm power based on big data and algorithm divide, which will affect the system of intention manifestation. Algorithm will not pose a fundamental challenge to the effectiveness system of legal acts, but it needs to be reinterpreted in the specific constitutive elements of mistake and fraud. When it comes to interpretation, a distinction needs to be made between different scenes where the user of algorithms is the actor or his counterpart in discussion. In the algorithmic society, consumers evolve to a new type of consumers: algorithmic consumers. In response to the inadequate algorithmic consumer protection in the existing private law framework, the mode of truly disclosure obligation in the Insurance Law should be adjusted, and the Consumer Law should reshape the rights and obligations between operators and consumers. In commercial and financial field, financial institutions using algorithms will have a higher duty of care. Multi-dimensional factors should be considered in algorithmic consumer protection. On the one hand, it is necessary to build a personal information protection system and take information fiduciary duty as corporate social responsibility. On the other hand, we should build the algorithmic transparency system, algorithmic audit system, algorithmic ethics and other comprehensive supervision systems to regulate algorithm. The essence of regulating algorithms through laws and rules is to regulate how algorithmic users (such as algorithmic businesses or platforms) use algorithms regularly and reasonably in the process of online transactions or operations. The present research on artificial intelligence should return to academic rationality and avoid blindly following the trend of fantastic jurisprudence. The core and key of algorithmic consumer protection in the algorithmic era lies in how to strike a balance between personal information protection, algorithm regulation and algorithmic enterprise development.

    参考文献
    相似文献
    引证文献
引用本文

刘颖.论算法与法律行为的关系:制度影响与法律回应[J].重庆大学学报社会科学版,2025,31(2):253-266. DOI:10.11835/j. issn.1008-5831. fx.2021.11.004

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2025-05-29
  • 出版日期:
文章二维码