Ask HN: How do you defend customer support AI agents from prompt injection?
1 theHolyTrynity 0 6/19/2025, 10:35:07 AM
We have built a customer support agent that does the following:
- retrieve data around company services from a RAG
- is connected to a few tools to escalate to humans and write support tickets
- has voice (11labs)
we did complete a POC but now we would like to audit this system to make sure it is safe against major attacks
we did try to follow the design patterns of deepmind, but wondering if there is any tool (preferably free / open source) that can red team our bot
No comments yet