Abstract: From the perspective of data sovereignty, algorithmic governance is challenged by the interplay between "data sovereignty" and "algorithmic power", the impact of nationalization of platforms and national platforms, and the core challenges of lack of prior governance solutions and time lag. In the process of tackling the challenges, various sovereign states have developed different governance approaches, including the proactive model of the United States, which focuses on multi-party accountability and risk response, the defensive model of the European Union, which is based on data protection with an emphasis on accountability, and the systematic attempt of China, which is based on the principle of overall national security. Drawing from international governance experiences, this paper proposes a development path for algorithmic governance from the perspective of data sovereignty of China,which includes the following key aspects: Firstly, regarding the governance principles, it emphasizes the need for a balance between development and security, advocating the "people-oriented" principle as the bedrocks of algorithmic governance. Secondly, regarding the governance agents, it stresses the importance of "governance-side" reform, calling for a collaborative, multi-stakehdder approach to mitigate algorithmic risk. Thirdly, regarding the govemance subjects, it highlights the criticality of "supply-side" reform, focusing on the enhancement of governance structures for both platforms and technologies. Lastly, regarding the governance methods, it promotes the advancement of a tiered classification system, which should be adeptly integrated with scenario-specific governance.

Key words: Algorithmic governance, Data sovereignty, Algorithmic power, Algorithmic risk, Platform responsibility, Overall national security