NOTE TO MODERATORS: I am fully aware that standard account issues belong in the Help Center. However, I am reporting a severe technical bug (an automated race condition causing a Global Mute) that is actively preventing the Help Center and Trust & Safety emails from functioning. Please read the timeline below before closing this topic, as the standard support channels are technically broken for my account.
Hello community,
I am posting this here out of pure desperation 
. I am trapped in a Kafkaesque nightmare with the automated support system. More than two years of conversations—ranging from highly valuable research data to personal life sharings and medical screenings—are currently held hostage by a bug/loop. If anyone has experienced this or if any OpenAI staff is reading, I urgently need a human override.
Here is the exact timeline of the paradox:
-
The False Positive Ban & Automated Wall
I am an AI professional and PhD candidate researching LLM security (specifically analyzing induced behavioral changes in LLMs such as sycophancy mitigation and theoretical propagation dynamics). My legitimate, sandboxed academic prompts triggered an automated “Cyber Abuse” flag. My account was completely deactivated, blocking my access to my conversations, the API, and freezing my prepaid API credits.
(Failed Login Request ID: 95d19a7-27bc-4f09-97f6-e77652b6227e)
Almost immediately, the automated system executed a “final decision” macro, effectively placing a global mute on my account. -
OpenAI Recognized Their Error (AFTER the block)
Some hours later, I managed to reach the Trust & Safety team regarding the warning. A human specialist reviewed my research logs, understood the academic context, and explicitly cleared me. I received an email stating verbatim:
“We have determined that we incorrectly issued a warning on your account.” (Ticket Ref: [C-ShSvecbdKQSk]) -
The “Ghosting” Bug
Here is the nightmare: even though Trust & Safety officially apologized and retracted the warning, the bot keeps my account deactivated because of that initial automated “final decision” macro.
When I reply to the human agent who cleared me to say, “Thank you, but I’m still locked out,” the system absorbs my emails. I am completely ghosted. I filed a parallel appeal (Ticket Ref: [C-c4wSK3oD5a9B]) days ago, but it’s sitting in a black hole.
The Devastating Impact:
This isn’t just about losing a subscription. I am locked out of over 2 years of daily workflows, code debugging, invaluable research conversations, and private personal history. My entire digital workflow has been wiped out by a race condition between a human agent and an automated bot.
It is incredibly demoralizing to do things by the book, be officially declared innocent by the security team, and still lose years of work because a ticketing system won’t let my emails through to the person who just cleared my name.
Ultimately, all I am asking is for OpenAI to respond at least one last time and clarify why my account remains locked. As a user, the overriding feeling is that I am simply trapped by an automation failure, and it seems clear that only a human supervisor can break this deadlock. Having been officially cleared of any wrongdoing by your own team, I believe I deserve at least a basic explanation for this profound contradiction.
If you read this far, an upvote for visibility would mean the world to me. If any OpenAI staff member can see the contradiction in these tickets and manually unlock my account, I would be forever grateful.
Thank you, community.