OpenAI employees are reviewing your conversations looking for ways to tweak & improve the program, so don't believe that things are more safe and non-judgmental than they actually are.