Accessibility Adjustments

Use these optional tools to adjust reading and display preferences. These tools cannot resolve every accessibility barrier. Please contact the website owner if you need assistance.

  • Text adjustments
  • Content scaling 100%
  • Font size 100%
  • Line height 100%
  • Letter spacing 100%
  • Colour adjustments
  • Orientation adjustments

Microsoft Autopilot enters Frontier preview, with US access and extra costs

Listen to this article

Microsoft Autopilot is rolling out through Frontier, bringing a persistent personal agent into Copilot. It is not a general release for every subscriber. Microsoft’s requirements limit the initial preview to selected organizational customers in the US, with administrator setup and usage charges beyond the Copilot license.

The distinction matters because several neighboring features are already generally available. Microsoft’s November 17 Home announcement puts Autopilot alongside Chat, Cowork, Projects and Automations, but describes Autopilot specifically as a Frontier rollout.

Who can turn on Autopilot?

Autopilot is off by default. An organization needs the applicable Copilot license, Frontier access, activation and an administrator spending policy. Autopilot usage is billed separately in Copilot Credits, not included with the license. Microsoft also lists Intune Plan 1 or Microsoft 365 E3 as a requirement for the execution environment.

This is an organizational preview, not a personal Microsoft account feature. A Copilot subscription alone is therefore not a reliable indicator that the agent will appear. Before planning a rollout, administrators need to check both eligibility and the budget that will fund ongoing work. Reaching the spending policy limit stops the agent from working.

What the agent can access

According to Microsoft’s overview, each person can have one Autopilot. It works through Copilot on the web or desktop and private Teams conversations, with its own cloud workspace and memory. It can continue recurring responsibilities through schedules and events rather than waiting for a new prompt each time.

The cloud workspace is not unrestricted access to the owner’s computer. Autopilot does not get local files and applications through that environment. Optional Edge access is a separate consideration, so readers should not equate a remote computer with control of everything on their laptop.

Home provides a place to chat with the agent and follow its activity across apps. That gives ongoing work a visible home, but users still need to check whether a task was completed, paused or left waiting for a decision.

Approval checks are not a guarantee

Microsoft’s application card describes automated checks that compare proposed actions with the request. Matching actions can proceed, while other actions require approval. Payments and sensitive browser actions always require approval under the documented controls.

The checker can make mistakes. That is a reason to set explicit boundaries before delegation, not to assume that any action taken automatically must have been appropriate. Memory is another deployment consideration because it cannot be turned off in this release.

The card identifies GPT-6.1 Sol through OpenAI as a subprocessor, with an Azure OpenAI-hosted GPT-5.6 Sol fallback when that OpenAI provider setting is disabled. Users do not get a model picker. Organizations evaluating the preview should review that routing alongside their own provider policies.

The documented compliance limits also matter: Autopilot conversations are not covered by eDiscovery, retention or legal hold. Sensitivity-label recognition does not mean generated files inherit those labels.

Start with a responsibility you can review

A useful first assignment is recurring work with a clear completion condition and an obvious point for human review. For example, ask for a weekly draft of unresolved project decisions without sending it to anyone. The practical question is not simply whether Autopilot can keep running. It is whether the organization has chosen what it may do, what it may spend and when a person must intervene.

Jordan Reid
Jordan Reid

Jordan Reid is focused on AI tools, agents, developer products, and the way technology changes everyday work. Jordan approaches a launch from the user’s side of the screen. What can it actually help someone finish? The voice is practical, conversational, and skeptical of products that turn a simple job into five new settings. Coverage follows coding assistants, creative software, browser agents, and the workflows around them, with attention to pricing, permissions, setup, and the human work that remains.

Leave a Reply

Your email address will not be published. Required fields are marked *

Gravatar profile