I’m reaching out because I’m a bit stuck. I’ve been running free AI-powered text-based tools and took a lot of precautions to stay within OpenAI’s usage policies. But despite all that, my account still got suspended, and I’m not sure what went wrong.
Here’s what I had in place to avoid any violations:
Rate Limiting: I set up IP-based rate limiting to prevent API abuse and stop spamming or overuse by any user.
Custom Explicit Word Filtering: I built my own filter to block inappropriate prompts before they even hit the OpenAI API.
Double Moderation: Every prompt was checked at two levels. First, it went through OpenAI’s Moderation API. On top of that, I had Gemini’s built-in safety settings activated to block anything related to harassment, violence, or other risky content.
With all these layers, I really thought things were covered. But somehow, my account still got flagged.
Has anyone else faced something like this? Could just a few bad prompts from users cause a suspension, even when double moderation is in place? I’m also wondering if anyone here has successfully appealed a suspension and can share how that went.
Any advice would be super helpful because I’m feeling stuck after putting so much effort into keeping things safe!
Gemini does not provide a separate moderation endpoint. If you are using Gemini’s “built-in safety settings” it means that you are using Gemini for your inference. Google does offer a moderation endpoint but it’s separated to Gemini.
The OpenAI Moderation endpoint is not meant to be a free service - it’s complementary. It’s expected that if you’re using it, you are using it with their inference.
Second, based on your previous thread I imagine that you didn’t also setup moderation for the output.
Your service is providing the most dangerous form of service: open, free access to inference. You should be doing everything and more to moderate the input and output.
Unfortunately you didn’t, and now the account is suspended. OpenAI’s support is notoriously difficult to communicate with, and sometimes can take weeks to respond. You should be sending proof that you are actually going to follow their moderation guidelines and hope that your account is re-instated
My advice would be to reach out to OpenAI and ask them to reinstate you, but first I would make sure you’re not letting anonymous users consume your service.