As described in Chap. 6 , an AIA may pose great threat to or even endanger human beings once the “three elements of uncontrollability” of AIAs appear. Therefore, it is important to take prevention and control measures in advance for AIAs. To improve the controllability of AIAs, Binxing Fang, Academician of the Chinese Academy of Engineering and editor-in-chief of this book, has proposed the concept of AIA safety hoop (hereinafter referred to as “safety hoop”) in a report titled “My Insight into AI Safety” made by him in the China International Software EXPO held on June 30, 2018. Here we take robot, the most representative AIA, as an example. Any robot should have a driving mechanism and a decision system. We can put the Safety Hoop between the driving mechanism and decision system, and the decision system must be driven only through the Safety Hoop. Just as a fuse will automatically blow when there’s an abnormal current to protect the safety of the power consumption system, the safety hoop will be activated when certain conditions are met, and then restrain the robot. By this way, measures can be taken to control a robot once it starts to lose control, and prevent it from becoming a destructive power and endangering human beings.