The principle "Universal Guidelines for Artificial Intelligence" has mentioned the topic "safety" in the following places: