Policies and laws regarding AI and related advanced technology are evolving in many countries. However, there has been little debate on the question of whether important decisions related to the use of AI technology should be decided at the administrative stage or in the parliament in the form of legislation. And there is a situation in which policy issues on AI are determined without much discussion or legislation. However, there are risks associated with utilization of AI, such as possible discrimination due to political decisions and there is a multi-layered problem associated with AI. This paper discusses the governance of risk control of AI should be decided with experts before making guidelines or legislation to balance the risk of AI technology utilization and transformation of current situation. To verify the regulation, the importance of keeping the transparency in the decision-making process should also be kept. This paper introduces the cases of Japan and EU.