URGENT: Data Loss Without Warning – Request for Immediate Recovery

To OpenAI Support,

I am a paying customer and have been using your services for years. I trusted your platform to retain critical data, and without any warning, a significant portion of my stored information was deleted or lost. This data was essential for an ongoing scientific research project on gastrointestinal lymphoma, and your system’s failure to notify me in advance has caused serious harm to my work and that of my colleagues. We last backed up the most recent data (that of a few hours observations) several hours to it being removed from your system.

This is completely unacceptable. If files become too large, I expect a clear warning before data is removed, allowing me to back it up. Instead, I was blindsided by this loss, despite being a paying subscriber who relies on your service.

I demand answers to the following immediately:

  1. Why was my data erased without warning?
  2. Is there any way to recover my lost data before it is permanently removed from your system?
  3. What are OpenAI’s actual data retention policies for paying users?
  4. Why was I not notified of any storage limits or potential deletions?

I need a direct response from your data team or developers regarding this issue, as well as any possible recovery options. If you cannot provide a solution, I will have no choice but to reconsider my subscription and explore AI services that respect their users’ data.

I expect a response immediately. This is not a minor issue—it has real consequences for my work. I will be monitoring this case closely and will escalate further if necessary.

Provide me with a resolution NOW.

If I do not receive a prompt resolution, I will be sharing this experience across multiple platforms to warn other paying customers that OpenAI deletes critical data without warning. This is especially concerning because we use OpenAI’s technology in our AI application, which is used by medical students, doctors, and other professionals who rely on accurate and persistent data. This kind of unexpected data loss could severely impact their work and trust in AI-driven solutions. If OpenAI cannot guarantee data integrity, it will force us to reconsider our reliance on your platform.

I will be posting about this issue on X (Twitter), Reddit, AI community forums, and other platforms so that others are aware of the risks of using OpenAI for critical work. I expect an immediate response and a solution to recover my lost data.

1 Like

It is possible that the data loss is related to the increased filtering of undesirable topics in ChatGPT recently.

his happened within the course of an hour. The subject matter was not undesirable. It was related to ongoing observations of case studies on steroid administration to diabetic patients with lymphoma and tracking the glucose trends and reactions to establish the correct insulin dosage. That was one aspect. There was nothing inappropriate. We use Open AI in our app developed for monitoring and giving possible outcomes and resolutions regarding patients in the veterinary sector since October 2024 and it has kept all of the data intact until yesterday evening when one of the students noticed the information was missing.

The data for an extremely important case study just disappeared within an hour. Also, we noticed it would not allow us to upload any files of data from PDFs, ie CBC reports necessary for the case. There are several veterinary patients whose lives could be adversely affected if we don’t get the data back. The data and Openai’s responses were extremely pertinent to the cases. We have some of the back data but are missing key observational information. We had noticed that the file upload was disabled prior to this occurring. However, we have not heard back from support at Open AI on what the resolution is. Our app tells us the data is there on its server but not visible on our side. However, not sure that is true because when we update cases, it acts like its the first time it has seen the data. It won’t allow us to upload the pdf backup tables to correct the data loss either.