I am curious about whether OpenAI will be able to see the prompts I enter. Suppose I am trying to use this API for processing sensitive data. Would this be a potential breach in security?
The model sees them and as far as I understand models they gather data to retrain once in a while because the schema of data’s changes with time, so where do they get that data? Might have a few sources where they get them, we can not say for sure its your data.
In your Settings of ChatGPT and also via emailing to OpenAI you can opt out of them using your data for model training. Just go to settings in your web application or your iOS application, or complete the following form to opt out: