So relieved to see someone else mention this, was starting to question whether I’m just being neurotic when little errors like this have me concerned about whether the underlying content is being thoroughly reviewed

Making an AI that detects violations of social norms itself sounds like a violation of social norms