Talk · AWS User Group Security Ecuador
What permissions do you give an AI agent on EKS and AWS?
Three questions, with live demos on a shop running on EKS: how to keep an agent from reading sensitive data or changing something nobody asked for, what stops it when someone gives it orders through a log line, and how to give it Bedrock models without anything leaving AWS.
Takeaways
If you're going to give an agent access
- Put the limits where it executes In the gateway, the session and the role. The prompt helps, but it doesn't protect.
- Redact before it reads What the model never sees, it can't leak, not even when tricked.
- Let it propose, and a person execute With their role, and with the server checking again.
- Find the outbound channel If there's no way to get something out, an injection wins nothing. In our case, it showed up in an image.