Microsoft Agent 365 Adds New Tools to Discover and Govern AI Agents

A centralized approach to discovering, governing, and securing AI agents across enterprise environments.

Cloud Computing

Key Takeaways:

  • Microsoft Agent 365 is now generally available with centralized AI agent management.
  • The new tools improve visibility and control over shadow AI and local agents.
  • The built-in security, policies, and integrations strengthen governance and protection.

Microsoft has announced the general availability of Agent 365, which first entered preview in November 2025. Alongside the release, the company introduced new capabilities designed to help organizations discover, monitor, and manage both sanctioned agents and shadow AI across their environments.

Microsoft Agent 365 is designed to give organizations a centralized way to monitor, manage, and secure AI agents across their environment. It is a framework with dashboards that show how AI agents are operating, including telemetry and alerts. It allows businesses to register AI agents with the Microsoft Entra registry, limit their access, ensure they integrate with Microsoft 365 apps, and protect against external and internal security threats.

With this release, Microsoft Agent 365 has added new features that allow organizations to find and manage OpenClaw agents on Windows devices. IT teams can use Microsoft Intune to detect where these agents are installed, monitor how they are being used across managed systems, and enforce policies to restrict or block unauthorized execution methods.

Microsoft plans to extend these discovery capabilities beyond OpenClaw. The upcoming updates will include support for other popular local agents such as Claude Code and GitHub Copilot CLI. This roadmap reflects a broader effort to gain visibility into the growing use of developer-focused and local AI tools.

Microsoft Agent 365 Adds New Tools to Discover and Govern AI Agents
Applying Microsoft Intune policies (Image Credit: Microsoft)

Policy-based governance and runtime protection

In June, Microsoft Agent 365 will introduce new policy-based controls that allow organizations to manage how AI agents operate within their environment. These controls enable teams to define clear rules governing what agents can access and the actions they are permitted to perform.

Additionally, Agent 365 will be able to detect and block suspicious or potentially harmful activities in real time. It will also generate detailed alerts with contextual information to help security teams investigate and respond effectively to incidents. These controls should help prevent misuse, data exposure, or unsafe automation.

Integration with Microsoft security ecosystem

Microsoft Agent 365 integrates closely with existing security tools to provide unified protection for AI agents. It leverages Microsoft Defender to monitor agent activity, identify potential threats, and support response actions. Moreover, Microsoft Entra enforces identity and network-based controls, such as limiting connections to risky external services.

Microsoft Purview also helps maintain data security and compliance by reducing the risk of oversharing sensitive information. These integrations enable organizations to use familiar tools and workflows to consistently secure AI agents across different environments.

Windows 365 for Agents

Microsoft has launched Windows 365 for Agents in public preview for US-based customers. It’s a cloud-based environment designed specifically for running AI agents in a secure and controlled way. It introduces a new type of Cloud PC that is tailored for agent workloads, which allows organizations to manage agents just like they manage employees’ devices. Agents can interact with applications, process tasks, and access resources while operating under defined policies, identities, and security controls.

This environment works alongside Agent 365 by providing a safe place where agents can execute their tasks while still being visible and governed through centralized management. It offers built-in security and integration into tools like Intune to help organizations ensure that agents operate within approved boundaries. This helps to reduce risks and makes it easier to deploy and scale AI solutions confidently in real-world business scenarios.

Microsoft Agent 365 is currently included with the Microsoft 365 E7 subscription or can be purchased separately for $15 per user per month. Meanwhile, to access the public preview of Windows 365 for Agents, customers need an Agent 365 license along with a Microsoft Intune license and an active Azure subscription.