The principle "A Proposed Model AI Governance Framework" has mentioned the topic "safety" in the following places: