Manage AI Agents Like Employees with Microsoft Agent 365
▼ Summary
– Microsoft is building an AI “agent factory” to enable businesses to create and manage their own AI agents, transitioning from its software origins.
– The company is launching Agent 365 as a control plane for businesses to manage AI agents similarly to how they manage human employees.
– Agent 365 provides dashboards with telemetry and alerts to securely deploy, monitor, and organize AI agents, preventing unexpected behavior.
– It integrates with Microsoft 365 apps, limits agent access via Microsoft Entra registry, and protects against security threats.
– Agent 365 will be available through Microsoft’s Frontier early-access program for IT admins to test AI agent adoption and management scenarios.
Businesses are on the verge of a major operational shift as Microsoft introduces Agent 365, a platform designed to let companies manage artificial intelligence agents with the same oversight and control they apply to human employees. This initiative represents a significant step in Microsoft’s evolution from a traditional software provider into a creator of autonomous AI systems capable of performing a wide array of tasks.
Jared Spataro, who serves as Microsoft’s chief marketing officer for AI at work, emphasizes the growing role of these digital workers. He points to IDC research forecasting a surge to 1.3 billion agents by 2028, illustrating their rapid integration into daily business functions. Spataro describes Agent 365 as the essential “control plane for agents,” effectively extending the trusted management infrastructure used for human staff to this new class of AI coworkers.
The core function of Agent 365 is to provide businesses with a secure framework for deploying and organizing AI agents, ensuring these digital entities operate predictably and safely. The system features comprehensive dashboards that display real-time operational telemetry and send alerts, giving administrators a clear window into agent activities. A key component is the integration with the Microsoft Entra registry, where each AI agent must be formally registered. This process allows companies to strictly limit what data and systems an agent can access, enforce integration with the broader suite of Microsoft 365 applications, and implement protections against both external and internal security threats.
Beyond its own ecosystem, Microsoft is ensuring Agent 365 supports a wide variety of AI agents developed by other industry leaders, including Adobe, Nvidia, ServiceNow, and Workday. The platform’s dashboards provide a unified view, enabling administrators to monitor the complex relationships and data flows between AI agents, human employees, and corporate information assets in real time.
Currently, Microsoft is offering early access to Agent 365 through its Frontier program, which serves as a testing ground for upcoming AI capabilities. This initial rollout is specifically tailored for IT administrators, allowing them to experiment with and refine the processes for adopting and overseeing AI agents within their organizational structures before a wider release.
(Source: The Verge)
