本文へ移動
KX3
AI・ローカルLLM / 約4分

Microsoft rebuilds Copilot around Home, Code and Autopilot for delegated work

Microsoft has introduced a new Copilot experience organized around Home, Code and Autopilot, combining Office work, natural-language app building and persistent agents that can keep working when users are away.

By admin@@
ノートパソコンで作業する人物のイメージ
Photo by Christian Velitchkov / Unsplash

KX3 News — U.S. / Global Edition

Microsoft introduced a redesigned Copilot on September 25, 2026, built around three new capabilities: Home, Code and Autopilot. The change pushes Copilot beyond a chat interface toward a broader work platform for creating, delegating and automating tasks.

Home combines Chat, Cowork and Office

Home becomes the main workspace for Chat and Cowork, while Office in Copilot brings Word, Excel and PowerPoint directly into the experience. Users can create or update real editable files and collaborate without constantly switching apps.

Code turns natural language into small business software

Code uses technology related to GitHub Copilot to build trackers, dashboards, automations and internal apps from natural-language requests. Microsoft says generated solutions can run in sandboxed environments and be hosted through Copilot Managed Runtime inside an organization’s Microsoft 365 environment.

Autopilot keeps working while users are away

Autopilot is a persistent cloud-hosted agent with its own identity, memory, computer and workspace. It can monitor channels, follow up on threads and continue recurring work without waiting for another prompt.

Availability

Home and Code are beginning to roll out through Microsoft’s Frontier program, while Autopilot expands to private preview at the end of September. Broader availability will follow for some capabilities. Microsoft did not announce a single new consumer price tied to this launch.

KX3 perspective

The larger shift is architectural: AI products are separating conversation, execution and persistent background work into distinct modes. That pattern is useful for smaller local or hybrid AI systems too, where reliability improves when each type of work has clear permissions and execution boundaries.

Primary source: Microsoft

日本語版を読む

Featured image: Photo by Christian Velitchkov / Unsplash.

Next read

関連記事

コメントを残す

メールアドレスが公開されることはありません。 ※ が付いている欄は必須項目です