With the development of “AI+” action, generative AI applications (AIGC) represented by ChatGPT, Sora, ERNIE Bot, etc. will gradually realize commercialization and become an important quality of new quality productivity. However, while AIGC improves people’s productivity and creativity, it also raises many data risks. Based on its technical principles, these risks can be divided into stages of illegal data acquisition risk, data security risk, and malicious content generation risk. Agile governance, as a new governance model, is a reflection and transcendence of traditional governance models such as exploratory, responsive, and centralized at that time. By embedding it into AIGC governance, it will have new connotations and characteristics in governance concepts, governance subjects, governance objects, governance tools, and other aspects. AIGC applies agile governance, which not only conforms to the trend of intelligent revolution and implements the requirements of technology for good, but also can overcome the difficulties of traditional governance. The guidance of national policy guidelines also creates a good legal environment for its application. Under the guidance of agile governance, the principle of risk prevention should be implemented and the data risk response mechanism should be strengthened. It is also necessary to strengthen the interpretation power of algorithms and build a data security guarantee system, improve the application generation mechanism, promote technology integration-based rule of law, effectively alleviate AIGC data risks, and empower the development of new quality productivity.