While receiving widespread attention, the fundamental issues in the research of artificial intelligence (AI) ethical risk prevention need to be probed and articulated. Although the relevant research is still at an initial stage with long-term tasks, it has a very important social significance. The strategic goals of AI ethical risk prevention research are clarifying ethical risks, taking flexible and effective measures, and providing ethical support for the development of AI technology. This paper argues that key tasks must be of the top priority, such as broadening the theoretical foundation, evaluating ethical risks, realizing ethical design and solving core ethical risks. Therefore, the research of AI ethical risk prevention can be substantially promoted.