These controls govern what an agent is allowed to do and how it is held to account. They carry the agent's intended action, separate tool access from actual tool use, and enforce least privilege at the hardware level, while ADR detects and responds when an agent is misbehaving or compromised. Together they form a defense-in-depth stack that both prevents and reacts to risky agent behavior.
Glossary topic
Agent least-privilege authorization and response
Terms in this topic
Track this in the live feed
See how this plays out in real AI security and governance developments.
Open the feed →