Windows 365 for Agents now in public preview: Run AI agents securely, at scale

AI agents are increasingly performing human-level work within enterprises, often interacting with applications through user interfaces rather than APIs. This reliance on visual interaction, where context and intent are conveyed through screens, presents a challenge for IT governance and management. Previously, AI agents ran on unsecured, ad-hoc infrastructure, creating gaps in security, auditing, and control. To address this, Microsoft is introducing Windows 365 for Agents in public preview, offering a secure, IT-managed Cloud PC environment for AI agents. This new service provides agents with dedicated, Intune-managed Cloud PCs, extending existing employee identity, security, and compliance models to AI workloads. Windows 365 for Agents integrates with Microsoft Agent 365, the control plane for defining agent behavior, policies, and managing permissions. Together, these solutions enable IT teams to confidently manage and scale AI agent deployments from experimentation to production. The platform offers enterprise-grade identity controls, unified device and policy management, and global scalability with geo-level data residency. This approach mirrors how employees work on managed devices, now extending the same trust model and IT controls to AI agents. Windows 365 for Agents provides the secure runtime environment, complementing Microsoft IQ's intelligence layer and Azure's cloud foundation.