News · November 17, 2026
Microsoft brings local coding agents under Agent 365 governance
At Ignite 2026, Microsoft extended Agent 365 and Entra to discover, identify and control AI agents running on developer machines, including GitHub Copilot CLI and Claude Code. For companies where engineers already use coding agents daily, agent governance is moving from policy documents into IT tooling.
What happened
- Organizations have registered more than 50 million agents in Microsoft Agent 365 since it launched in May 2026, Microsoft said in its Ignite 2026 update on Agent 365, published November 17, the first day of the conference.
- Microsoft said local agent discovery in Microsoft Defender is now generally available, and that Defender runtime protection for local agents is generally available on Windows; network-level discovery in Global Secure Access is in preview.
- Entra Agent ID, which gives agents their own identities, now extends in preview to discovered local agents, starting with GitHub Copilot CLI and Claude Code, according to Microsoft; Virtualization Review reported that the preview covers 60 local agent types, including OpenAI Codex.
- Microsoft also announced in preview Intune controls for local agents, Global Secure Access controls for agents, and Purview data loss prevention for Microsoft agents and key local agents.
- Virtualization Review reported that automated lifecycle management for agent identities is generally available, with a default policy that disables agent identities inactive for 60 days.
Why it matters
Coding agents have mostly lived outside central IT: installed by individual engineers, running with the developer's own permissions and touching source code, secrets and production consoles. Microsoft's move signals that large customers will soon expect the same inventory, identity and data-loss controls for these tools that they already apply to laptops and SaaS apps. Most of the new controls are still in preview, so this is a direction to plan for rather than a switch to flip today.
For engineering leaders, the practical step is to write down which agents your team actually runs, on which machines and with which credentials, before a security review asks. Give agents scoped tokens instead of personal admin keys, keep secrets out of repositories they can read, and log what they change. That work pays off whatever governance stack you end up using.
For companies working with outsourced or dedicated teams, expect agent usage to become a contract and onboarding topic: which coding agents are allowed on client code, under whose identity they run and how their access is revoked when a project ends. Agreeing on this up front is far easier than retrofitting it after an audit.
Sources
- Microsoft Community Hub (Agent 365 Blog): What's new in Microsoft Agent 365: Ignite 2026
- Virtualization Review: Microsoft Ignite 2026: Top Azure Announcements for Enterprise IT
This post is CMM’s own summary and commentary on publicly reported news. Facts are attributed to the sources linked above; please read them for full details. CMM is not affiliated with these publishers or with the companies mentioned, and trademarks belong to their owners. Prepared with AI assistance. Corrections or requests: [email protected]. See our editorial policy.
Want help with this?
Tell us what you need, and we will suggest the right team model on a short call.
Talk to us