Abstract: In a centralised political system, smart communities function both as an approach to enhancing the efficiency of grassroots governance and as a tool for maintaining social order and political stability. This article develops an analytical framework of the "technology–power" dynamics, categorising governance models into four types: technology control, technology empowerment, traditional control and grassroots autonomy. Drawing on case studies from China, the findings indicate that while the technology control model increases governance efficiency, it restricts civic autonomy and raises privacy concerns. Conversely, the technology empowerment model fosters democratic consultation, but its implementation is constrained by political sensitivities. The traditional control model relies on administrative mobilisation for policy implementation but lacks adaptability. The grassroots autonomy model implies community resilience but faces challenges of sustainability. This article provides a theoretical framework for understanding smart community governance in centralised systems and offers empirical insights from a non-Western context, thereby enriching the research and literature on global smart governance.