ITHome News on October 8: Microsoft held its Surface new product launch event today, where it demonstrated the future direction of Autopilot. According to its vision, this agent will enable Windows AI to further achieve autonomous task execution.
Under Microsoft's vision, Autopilot will no longer just process information based on user instructions, but will be able to understand goals, plan steps, and call upon the Windows system and various applications to complete tasks. Users can describe their needs in natural language, whilethe AI is responsible for finding information, manipulating files, configuring systems, or executing workflows，the user is responsible for confirming or intervening at key junctures。
ITHome has learned that Autopilot's positioning echoes Microsoft's current 'hybrid intelligence' philosophy: humans set goals and retain final control, while the AI handles information processing and concrete execution; meanwhile, it combines local device and cloud capabilities to strike a balance between response speed, privacy protection, and task complexity.
However, these highly autonomous agents inevitably carry security risks. In the future, Microsoft will need clear authorization mechanisms, second confirmation before execution, and traceability mechanisms to ensure Autopilot never slips out of the user's hands.