In recent years, the rapid development and widespread application of generative artificial intelligence (GAI) technology have, on the one hand, brought opportunities for transformation and upgrading to socio-economic development, while on the other hand, new risks are also emerging rapidly. In this digital transformation environment, accurately identifying the safety risk points of generative AI is a crucial safeguard for the safe governance of generative AI. This paper employs a procedural grounded analysis of policy texts, conducting procedural coding on the content of 53 policy documents to construct the “ECA”model, which encompasses three core categories, 11 main categories, 31 subcategories, and 92 basic categories. Based on the four stages of digital transformation, this paper identifies and analyzes the risks faced by GenAI in each stage focusing on the characteristics of cocooning, pupation, metamorphosis, and ecologicalization in the digital transformation environment. It proposes a comprehensive dynamic governance strategy encompassing risk perception, early warning, monitoring, risk tiered management, and response.