The principle "OpenAI Charter" has mentioned the topic "safety" in the following places: