OpenAI Handling of Client Code

Hi - our firm is testing out a suite of services where openai codex may be used for part of the process.

Does OpenAI keep copies of code that are submitted to the Codex engine? The specific point we want to understand is whether a customer of our platform could be assured that proprietary digital assets would be treated confidentially, and no copies of code, for example, are kept by openai (or at least that there is an opt-out option where needed)

Thank you


Just for reference, I posted the reply below that I received through the Chat. I know there were a couple of other folks interested in the answer on the question

Hi there!

Thanks for reaching out; you can review our Privacy Policy for more details about how we handle submissions of your prompts. If you’d like to opt out of data training, you can follow the steps here.