Hi, i'm looking to use AssistantsAPI for my support service in my discord server. Since my budget is quite low, I'm looking into the "Content Sharing Agreement". If I put up a notice and filter certain keywords like "address", "medical info" etc before running GPT, but a user still finds a way to go around it and input something like medical information, am I then still responsible for that person's actions? / how could I prevent this? Thank you.
#Content Sharing Agreement Question
marble basin
I also just looked into the moderation API. If I filter every input through that first, should that prevent abuse / me from being liable for the end user’s input?
vital sable
I also just looked into the moderation API. If I filter every input through that...
that and following the other best safety practices in docs like userid etc. will not result in instant/perm account suspension but warning mails first if they occur often
vital sable
Hi, i'm looking to use AssistantsAPI for my support service in my discord server...
there is no flawless way to cover all cases and the sharing is not really mean for PII or other similar usage but safely more like data processing, automation, etc.
marble basin
Alright, thank you very much!!
marble basin
Hi, I just had one more question. Does the omni-moderation also apply for the data sharing? It's free of charge to use but does it also store data or give the inputs to OpenAI when you have the content sharing option enabled (or is that only for GPT models)?