TC260 Security Guidance for Agent Deployment and Use

Published July 2026. Addresses the security discipline expected of organisations deploying and operating AI agents, covering permissions, identity, monitoring and containment of agent actions within the deploying organisation's infrastructure.