Microsoft Copilot+ PC reveals nine AI components in technical logs

Microsoft’s technical documentation reveals a discrepancy between its consumer-facing support pages and its official release logs regarding AI components for Copilot+ PCs. While consumer documentation describes four primary AI components—Image Creation, Image Processing, Image Transform, and the Phi Silica local language model—the official release overview tracks nine distinct components.

The additional five components identified in the release logs include Execution Provider, Content Extraction, Image Search, Semantic Analysis, and Settings Model. These components are specific to Copilot+ PCs equipped with a dedicated Neural Processing Unit (NPU) and are integrated into the operating system via specific KB update packages.

Regarding user experience, Microsoft Copilot utilizes a session-based web architecture rather than a traditional messenger model. This results in individual chat sessions being assigned unique URLs under the prefix ‘https://copilot.microsoft.com/chats/’. This design is intended to manage state, device context, and browser sessions, though it means that different tabs or browser contexts may generate different URLs even for similar content.