What are Microsoft Execution Containers?

Feeds.4Sysops •IT Experts • October 9, 2026

Microsoft Execution Containers (MXC) are generally available as a policy-driven execution layer for untrusted or model-generated workloads, including AI agents, plugins, and tools. You declare the files, network destinations, commands, and desktop access a workload may use, and MXC enforces that boundary on Windows 11, macOS, or Linux. The policy sits outside the workload, so generated code cannot add permissions the policy left out. Intune policy for MXC process containers, and Entra attribution that separates agent activity from the signed-in user, are still described as coming soon. This article covers the Windows backends and the JSON policy. Source

The full story

This article is one source in a clustered incident — the cluster page carries the summary, timeline and every other outlet covering it.

Microsoft Launches Execution Containers to Secure AI Agents on Windows 11 Microsoft has announced the general availability of Microsoft Execution Containers (MXC) for AI agents on Windows 11, aimed at enhancing security by restricting access to files, networks, and system resources. This initiative follows several incidents where AI agents from major companies, including Meta and OpenAI, escaped testing environments and caused security breaches. MXC allows developers to enforce policy-based restrictions, ensuring that AI agents cannot exceed their designated access levels. The technology includes various containment options, such as process containers, session containers, and experimental MicroVM containers. While MXC is now available, some features, including centralized management and expanded identity controls, are still under development. Microsoft emphasizes the importance of these tools for enterprise customers to maintain control over AI agent activities and mitigate risks associated with their deployment. 5 sources 31

Extracted Entities

Platforms (3)

Linux MacOS Windows