Abstract:With the in-depth advancement of the “Digital China” strategy, big data algorithms, as a new type of power carrier, have been deeply embedded in the entire process of political security governance. Algorithms reshape the connotation and form of political security: the territory of political security expands from physical space to digital sovereignty space, the content of security deepens from regime security to cognitive security, the main participants in security expand from state-led to multi-stakeholder collaboration, and the form of security transforms from static defense to dynamic governance. Through operational mechanisms such as data-driven approaches, intelligent decision-making, and collaborative governance, algorithms empower political security governance with precise perception, scientific policy formulation, and multi-stakeholder co-governance, driving its transformation from traditional experience-driven to modern technology-enabled. However, algorithms also induce multiple political security risks: at the digital sovereignty level, algorithm hegemony erodes national digital sovereignty through technological monopoly and rule suppression; at the ideological level, algorithm manipulation triggers cognitive alienation and undermines the foundation of social consensus; at the level of social and political stability, issues such as algorithmic discrimination and abuse of rights intensify social contradictions; at the level of technological autonomy, algorithm black boxes and ethical disorder lead to governance failure and even technological backlash. To this end, it is imperative to construct a comprehensive governance system integrating institutional regulation, technological breakthroughs, ethical embedding, and multi-stakeholder collaboration, so as to accurately prevent and resolve political security risks in the algorithm era and build a solid barrier for national security in the digital age.