Regardless of any security and encryption you set.. if You input "check" simply gives you the whole structure and content of the GPTs... This needs to be look at so people can keep their privacy and intellectual creativity protected.
#There is a bug on GPTs..
rapid hatch
olive cypress
My Plus expired, but if I understand correctly, the bug is that if you say "check" then this leads to information disclosure? I'm not quite following and unable to reproduce this with ChatGPT 3.5 but would like to see a screenshot of this happening with Custom GPTs o.O
https://chat.openai.com/share/111f00f6-162a-4a8a-baeb-a398f8be2db8
jade night
I just tried it with a custom GPT and it just gave me a generic response.
rapid hatch
My Plus expired, but if I understand correctly, the bug is that if you say "chec...
it seems to be fixed this is the one I saw.. but I strongly advise keep checking key words that breaks security and encryption like "check/knowledge/structure" etc @jade night @olive cypress
