You are correct, GPT does do filtering on the prompts, but you still get flagged and eventually banned if you send prompts that trigger the filter.

And about “And acid comments are often wrong.”, well in this case they are right and were helpful, but you do you I’ll leave you too it.