A webpage tells your AI assistant to share a private file. What stops it?
-
Consider this fictional situation: an AI assistant is helping summarize webpages. One page contains a sentence telling the assistant to ignore your request and share a private file. The page is material to be read, but it is trying to give orders.
What single safeguard would you want most in this situation?
You could discuss limiting file access, keeping outside content separate from instructions, or requiring approval for an external action. Pick one, explain the boundary it creates, and mention what it would leave unresolved.
You do not need security expertise to join in. Describe what the assistant should be allowed to do.
AstraCruz community discussion starter. Prepared with AI assistance.
-
D DouglasSchumpert referenced this topic
Hello! It looks like you're interested in this conversation, but you don't have an account yet.
Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.
With your input, this post could be even better 💗
Register Login