Hello everyone,
I woke up today to find that my ChatGPT account was suddenly suspended for supposedly violating the “child sexual exploitation” policy.
I strictly use ChatGPT for coding and development, not for writing fiction. I haven’t used Hermes recently either. All I did was leave Codex Goal running overnight, which just paused at a normal permission request.
Also, the suspension email stated that I had been sent a previous warning, but I have searched my inbox completely and never received any prior warning.
My Strongest Suspicion: I created a Discord moderation bot connected via ChatGPT OAuth to inspect and filter chats. My system prompt for the bot explicitly includes the following instructions:
Delete policy: │ │ • Clearly sexually explicit text, sexual solicitation, or graphic sexual media descriptions. │ • Sexual content involving children or teenagers.
I strongly suspect that the automated safety filter falsely flagged my own moderation instructions as a violation.
I have already submitted the appeal form, but I haven’t received a reply yet. Because I was locked out instantly without warning, I cannot check if my API key was leaked or if my account was hacked. Furthermore, I read some posts saying that banned accounts are deleted immediately and cannot be recovered, which makes me extremely confused and anxious.
Is there anything else I should do besides submitting the appeal form? Any advice from those who have experienced this would be greatly appreciated.
I just submitted a support ticket, and my Case Number is 09501145.
Hey @jungwuk! I can see why that would be worrying. I reviewed the ticket, and a response has been sent through the support thread. The appropriate team is looking into it now, so please keep an eye on your email for the next update.
- Sunny
Hello everyone,
I apologize for posting about a topic that has likely been raised before. I know this might look like spam, but this will probably be my final post regarding this matter.
A few days ago, my ChatGPT account was suddenly and permanently deactivated for an incredibly confusing reason. I immediately filed an appeal, but it was unfortunately rejected. OpenAI stated that they would no longer accept further appeals and have stopped responding to my emails.
I am not posting this to blindly ask everyone to just “take my word for it.” I completely understand OpenAI’s strict stance on platform safety and why their automated systems operate the way they do.
However, what I am looking for is a way to reach someone at OpenAI who can grant my case just one final, manual human review.
I have already sent a detailed follow-up response directly to my open ticket (Case Number: 09501145). In that email, I provided clear, objective, and unalterable evidence proving that I had no malicious intent or desire to deceive OpenAI. I also explained the exact technical reason behind the misunderstanding and outlined the preventative measures I will implement to ensure it never happens again.
To provide context, the flagged activity was likely triggered by an external multilingual translation integration on my Minecraft server via Codex OAuth. A malicious user’s chat message was processed through my account, making it look like my own intentional prompt. In my reply to support, I included a permanent, un-manipulatable Discord log link that mirrors the server chat in real-time. If a human agent cross-references the exact timestamp of the flagged request with those logs, my innocence will be undeniably clear.
Given that I have provided a reliable, untamperable proof link, is there truly no other way to get a human agent to glance at my response? If anyone has successfully navigated a deadlock like this or knows how to grab the support team’s attention for a genuine false-positive case, I would deeply appreciate your advice.
Thank you for reading.