There MUST be some extra risk assessment based on your account activity and I have had my account since gpt 3 and used it insane amounts, almost every day for years. I see people all over the internet getting denied for hypothetical "cyber" questions. I use it for actual cyber and code security all the time, and only got a quick refusal once that told me about signing up for "daybreak" or whatever when I had codex astra do an adversarial code security review, and I just restarted it and it did it anyway.
Date: September 21st, 2026 10:01 PM Author: Brilliant Exciting Stage International Law Enforcement Agency
I take it back. A couple of years ago I tried to have a conversation with ChatGPT about the episode of My So-Called Life where Jordan Catelano wants to take Angela to that vacant house to have sex and Angela, a virgin, keeps looking for someone to tell her she's not allowed to have sex so she doesn't have to take responsibility for telling Jordan she's not ready. ChatGPT deleted the conversation. I get why, but it's dumb that you're not allowed to talk about something that was considered appropriate for an after-school special 30 years ago.
Sometimes for cyber related requests I get the "hold on this request needs additional processing, if you want you can downgrade the model, but it might not be as capable of handling complex requests", but it always eventually goes through