It's not a bug when invoking against the GPT model (not ChatGPT) directly. Such a model needs to be able to understand and produce that content. The "what you do with it afterwards" is where it needs to be examined.

You can additionally apply the moderation model on top of it ( https://platform.openai.com/docs/models/moderation and https://platform.openai.com/docs/api-reference/moderations )

Note that these are separate services and have different goals.