← sources

Received a "Cyber Abuse" warning, and it was Astra's doing

source: reddit (reddit:OpenAI)published 2026-11-09 02:02ingested 2026-11-10 02:39

retained text

I added a chat box for customers to a website, and asked Astra to make sure it was secure, and that we were not allowing non-related chats in our support chatbot.

Well I get a warning about an hour later from OpenAI accusing me of "Cyber Abuse".

Turns out Astra went crazy and started trying to jailbreak my chatbot.

Mind you, I did not ask for this kind of testing. Astra went above the call here. I didn't even ask it to test. I just asked to to make sure it was secure.

I have an appeal in with them now, but jeesh, their own model caused this, and I'm being blamed.

  
Please, do better OpenAI.

document fetched 2026-11-10 00:27 · canonical https://reddit.com/r/OpenAI/comments/1xrfi1j/received_a_cyber_abuse_warning_and_it_was_astras