I don't think I get what you are referring to with malicious intent, but OpenAI has some "Safety best practices" [1] that don't seem bad. I'm also sure that the

I don't think I get what you are referring to with malicious intent, but OpenAI has some "Safety best practices" [1] that don't seem bad. I'm also sure that they will improve over time. [1] https://beta.openai.com/docs/usage-policies/safety-best-practices https://beta.openai.com/docs/usage-policies/safety-best-prac...