Microsoft Rebuilds Copilot Around Agents: Meet Autopilot, Code, and the New Home (September 2026)


On September 25, 2026, Microsoft announced its biggest redesign of Copilot since the product launched. The workplace AI is being rebuilt around three pillars — Home, Code, and Autopilot — and the headline is Autopilot: a persistent, autonomous AI agent that lives inside your organization's Microsoft 365 environment and keeps working when you're not looking. Here is what was announced, when it arrives, and what to watch before your company turns it on.

The Three Pillars

  • Home merges the familiar Chat with Cowork into a single front door for everyday questions and document work.
  • Code lets employees describe an app, dashboard, or widget in plain language and have Copilot generate it.
  • Autopilot is a standing autonomous agent with its own identity, memory, and workspace inside Microsoft 365.

What Autopilot Actually Does

Autopilot is the renamed, expanded version of what Microsoft previously called Scout. Instead of answering one prompt at a time, it monitors projects, coordinates with employees across Teams, Outlook, and documents, handles recurring tasks, and resumes work after long gaps. In Microsoft's demo, an agent named Dot prepared for Black Friday by pulling together email threads, Teams discussions, and inventory data, then flagging and resolving a shipment problem.

CEO Satya Nadella framed the guardrails this way: every agent has to have an identity, and everything it does needs to be observable.

Code: Apps as Easy as Documents

With Code, you describe the tool you need and Copilot builds it. The generated apps run on the new Copilot Managed Runtime, hosted inside your organization's Microsoft 365 boundary rather than on some external service. The code stays editable with Git version control, and an SDK will connect third-party tools. Microsoft's pitch: sharing an app becomes "as simple as saving and sharing a Word document."

When You Can Use It

  • Home: rolling out to Frontier program participants over the coming weeks
  • Code: reaches Frontier at the end of September; broader access and Microsoft 365 Premium/Pro availability planned for later in 2026
  • Autopilot: private preview expands at month's end; the managed runtime enters public preview

For most users, this is a staged rollout — what appears in your own tenant is the real signal, not the keynote.

Pricing: Subscription Plus Credits

The enterprise subscription ($30 per user per month) covers everyday use, but the heavier features — Cowork, Code, and Autopilot — consume per-use credits on top. Microsoft has not published the consumption rates yet. Earlier reporting around Cowork suggested light tasks cost roughly $0.70–$2 in credits, medium ones $4–$6, and heavy jobs $15 or more, with annual commitments discounted 5–20% — treat those as estimates until official rates land. The two-layer model is spreading across the industry for a simple reason: an agent that works in the background for hours burns real compute, and flat pricing can't absorb it.

What to Watch Before Adopting

  • Demos show intended workflows; independent reliability across real deployments is unproven so far.
  • Detailed approval policies — which agent actions need a human sign-off — remain undocumented.
  • Credit consumption rates are unspecified, which makes cost forecasting impossible for now.

FAQ

Is this for personal Copilot users too?

The announcement centers on business Microsoft 365 environments. Consumer availability beyond the planned Premium tiers has not been detailed.

Does Autopilot act without permission?

Microsoft emphasizes that agents operate inside the organization's boundary with observable actions, but the fine-grained approval rules haven't been published — ask before enabling it.

Is the old Copilot chat going away?

No. Chat continues, folded into the new Home experience alongside Cowork.

Related on AI Learning Lab: ChatGPT, Claude, Gemini, or Copilot on Your Desktop? · What Is Agentic AI? · Amazon Opens Its Seller Tools to Outside AI Agents


Comments

Post a Comment

Popular posts from this blog

AI Search vs Traditional Search: How to Use Each in 2026

Image
Something changed in how people find things, and the numbers are unusually clear about it. According to a SparkToro analysis of Similarweb clickstream data published in June 2026, about 68% of U.S. Google searches between January and April 2026 ended without a click — up from roughly 60% in 2024. Answers are arriving without anyone visiting a page. That makes "which kind of search should I use?" a question worth having a real answer to. What the data actually says Two figures from that analysis matter most. AI Overviews now appear on more than 20% of Google searches, and click-through rates fall by roughly 60% when they do . The behavior shift is not that people stopped searching — it is that a search increasingly ends where it used to begin. For anyone doing research, the tradeoff is speed against provenance. You get an answer faster, and you can see less about where it came from. When AI search is the right tool You do not know the vocabulary yet — "the clause that l...
Read more

Free vs Paid AI Tools: When Is Upgrading Actually Worth It?

Image
Every AI assistant now has a free tier that's genuinely capable, and a paid tier that costs roughly the price of a couple of lunches per month. So the question isn't "is AI worth paying for" in the abstract — it's whether your specific usage crosses the line where paying changes anything . This article gives you that line. What you actually get when you pay Pricing and features shift constantly, so ignore the marketing lists and look at the three differences that persist across providers: Priority access. Free tiers get throttled or queued at busy hours. If your work happens during peak times, this is the difference between a tool you rely on and a tool that sometimes isn't there. The current top model. Free tiers usually route you to smaller or older models, with limited access to the best one. For simple tasks this doesn't matter. For reasoning-heavy work, it's the whole ballgame. Data handling. This is the one most people overlook. Paid and busin...
Read more

What Is a Small Language Model (SLM)? On-Device AI Explained

Image
Most of the attention in AI goes to models that get bigger. A quieter trend runs the other way: models small enough to run on a laptop, a phone, or a server in your own building. These are usually called small language models, or SLMs, and they are showing up inside products you already use — often without being announced. Here is what they are and when the small option is the better one. What counts as "small" There is no official threshold, but the working definition in most technical writing puts SLMs under roughly 10 billion parameters, with the common range falling between about 1 billion and 7 billion. Frequently cited examples include Microsoft's Phi-3 Mini at 3.8 billion parameters, Llama 3.2 3B, and Mistral 7B. The contrast is with frontier models, which run to hundreds of billions of parameters and, in the largest cases, over a trillion. Parameters are the adjustable values a model learns during training — loosely, a measure of how much the model can store. Fewe...
Read more