As digital government systems evolve, increasing complexity in information interactions has challenged traditional hierarchical governance models, which often struggle in dynamic and cross-sectoral contexts. This study aims to identify the structural patterns of government information interaction and to develop a testable adaptive governance approach that supports sustainable digital government evolution. Drawing on IT alignment theory and complex network analysis, this study reconceptualizes digital government as a complex adaptive system and reveals the heavy-tailed distribution, structural stability, and self-organizing tendencies of government information networks. Building on these findings, the study develops and operationalizes a self-organizing adaptive governance framework—featuring fractal design, dynamic alignment, and layered modular coordination—into 11 governance rules. By shifting the focus from static alignment to adaptive structural coordination, this research advances a new pathway for the sustainable and resilient evolution of digital government systems.