No, we have never done that. What made you think we were? What is more likely coming soon though is that we will have to start sending a "user id" hash to the likes of openAI so we can block certain users when openAI deems the user is doing bad stuff. This has become a concern in the last couple of ...