Microsoft's MXC puts AI agents in a policy-driven box
Mxc: Microsoft Execution Containers version 1.0.0
Microsoft Execution Containers (MXC) is now generally available, giving developers and IT a policy-driven execution layer that enforces what AI agents can access at runtime. It offers process, session, WSL, and MicroVM containers across Windows, macOS, and Linux, with a unified JSON schema and SDK. Microsoft Entra and Intune integration will add agent identity and management, while partners like GitHub Copilot and OpenAI Codex already support MXC.
An agent cannot be its own security authority. It must run within a boundary defined by the developer or organization and enforced independently of the agent itself.