Microsoft expands agent security with dynamic permissions and Guardian Agents
🤖 Microsoft announced an expansion of Project Perception to protect AI agents as they work across enterprise systems.
🤖 Discovery now covers 60 local agent types, while Entra Agent ID gains support for additional agent patterns.
🛡️ New Dynamic Permissions scope access to a task and its duration; Dynamic Agent Risk and Dynamic Conditional Access adjust protection as behavior changes.
💡 Guardian Agents and Playbooks coordinate responses within human-defined policies.
🤖 For teams building or securing customer agents, the announcement connects identity, least privilege and continuous containment.
📌 The post does not establish general availability or licensing for every capability.
Partner briefing
- What it is
An expansion of Microsoft's security architecture for enterprise AI agents.
- What’s changing
Broader discovery, task-scoped permissions, adaptive risk controls and Guardian Agents.
- Why it matters
Protection can narrow access or contain execution as agent risk changes.
- Who it affects
Enterprise identity, security and AI development teams.