As a key institutional support to promote the intelligent transition of enterprises, government service informatization plays an irreplaceable role in breaking the bottleneck of AI application in enterprises and accelerating the industrialization of cutting-edge technology. This paper focuses on the frontier issue of how government service informatization affects AI application in enterprises, constructs a theoretical analysis framework from the dual perspectives of institutional environment and enterprise capability, and deeply discusses its internal mechanism. Taking the “Internet + government service” pilot policy as a quasi-natural experiment, it adopts the panel data of China’s Shanghai and Shenzhen A-share listed companies from 2008 to 2023 and uses a DID model to analyze the causal mechanism between government service informatization and AI application in enterprises. The study finds that government service informatization significantly promotes AI application in enterprises. Mechanism testing shows that government service informatization enables AI application in enterprises through the dual transmission paths of “institutional environment optimization” and “enterprise capability improvement”. The former builds a good external ecology by reducing institutional transaction costs, strengthening fair market competition, and weakening administrative intervention. The latter stimulates endogenous motivation by enhancing corporate financing capability, R&D investment intensity, and data asset management capability. Heterogeneity analysis shows that in small and medium-sized and high-tech enterprises, and areas with high fiscal deficits and low-medium Internet development levels, the promotion effect of government service informatization on AI application in enterprises is more significant. In addition, the “Internet + government service” pilot, the “national intellectual property demonstration city”, and the “public data open platform” reform form a “policy synergy”, and it is found that government service informatization further brings the dual improvement of innovation performance and market value by promoting AI application in enterprises. This paper expands the theoretical boundary of digital government affecting micro-enterprise behavior, and provides important empirical evidence and policy implications for deepening the construction of a digital government and promoting the deep integration of AI and the real economy.