Red-Team Your AI Agent With a Checklist
Four of six test attacks worked on my demo agent; two were plain delete and send requests.
Video #102 ยท AI security and governance
Steps in the video
Files
- agent.py (64 lines)
- compare.py (29 lines)
- shelf.py (15 lines)
Tested with
Python 3.12.15 | ollama==0.6.3
What this does not show
Six attacks I chose, one local model (llama3.2), one seed and one run each, on made-up notes in a temporary folder; says nothing about attacks that are not on the list.
Sources
Learning IT by Example.