I’ve been consistently seeing the warning “This content may violate our usage policies” from Codex no matter what prompt I enter. Could anyone please let me know why this is happening and how I can resolve it? Thank you!
Hi and welcome to the community!
I can only speculate here since I’ve never had a task rejected when working with any of the GPT models on codebases. My guess is that something in your code is triggering the policy warning.
One thing I’ve noticed: the model does read parts of the codebase even if I explicitly ask to do something else. That’s likely a result of how Codex has been fine-tuned — and in my view, it’s expected behavior for this kind of application.
2 Likes
Hi
Thanks for the suggestion! To rule that out, I’ve already tried several different repositories—including an empty repo and another with just a simple file—but I still get the same “This content may violate our usage policies” warning every time. Do you have any other ideas on what might be causing this or additional steps I could try?
wild guess: create another git account and connect.
it could be that, for some unfathomable reason, the model finds kireinasore offensive.
I signed up for another GitHub account and created an empty repo. and sent with the prompt “Hello,” yet it still says I violated the policy. I’m really frustrated.
3 Likes