生成式人工智能视域下个人信息保护的风险路径与算法治理
Risk Paths and Algorithmic Governance of Personal Information Protection in the Perspective of Generative Artificial Intelligence

生成式人工智能的涌现使得个人信息保护面临着日益严峻的风险。本文探讨了生成式人工智能技术的算法逻辑以及在个人信息保护中产生的风险，分析了个人意思自治空间被压缩、个人信息频繁泄露、数据合规外延扩大、主体权利价值被削弱等风险路径。算法作为规则的正当性仍需深度论证，但其在个人信息保护制度中已初见雏形，完善个人信息保护制度中的算法规则对保护主体信息权益有重要作用。在算法治理进路上，我们以算法可解释性为契机重构算法二元治理进路，打开生成式人工智能场景下算法治理的切口，以此实现个人信息保护和流通的平衡。
The emergence of Generative Artificial Intelligence (AI) has put the protection of personal information at increasing risk. This paper discusses the algorithmic logic of generative AI technology and the risks arising from personal information protection, analyzes the risk paths of compression of the space of personal autonomy, frequent leakage of personal information, expansion of the extension of data compliance, and weakening of the value of the subject’s rights. The legitimacy of algorithms as rules still needs to be demonstrated in depth, but it has taken shape in the personal information protection system, and improving the algorithmic rules in the personal information protection system plays an important role in protecting the subject’s information rights and interests. On the way of algorithmic governance, we take interpretability of algorithms as an opportunity to reconstruct the algorithmic binary governance approach, and to open the cut of algorithmic governance in generative artificial intelligence scenarios, so as to realize the balance between the protection and circulation of personal information.