If I am creating a simple application for my own personal use, which makes use of the OpenAI API, am I required to adhere to the OpenAI Usage Policies and enforce moderation on interactions?
To be clear - this is not a program I intend to let others use; the only way that happens is if someone breaks into my computer and steals it. So if I’m not enforcing adherence to the policies through use of the moderation endpoint when I am the only person using the program accesing the endpoint, am I risking my key getting suspended/terminated?
Thanks.
I imagine that if it is having to send a request to OpenAI and it contains content that is against their policies, then yes that will likely pose a risk to your standing. I do think that OpenAI should offer a “guardrails off” version that also releases liability as well. Focus on ensuring the accuracy of the model, not restricting access to it. I want to do something similar to you in which I make a personal caretaker app and I am extremely aggravated at the idea of having to censor myself in such a way, but until such time as OpenAI makes a change in policy or another option becomes available I suggest implementing the moderation and add a buffer in between your input and it going to OpenAI so that you can pass off moderated content to another language model that allows for explicit content as an exception, with another exception for there being no such model and getting censored by the mod.