At the Microsoft Build2025 conference, GitHub and its parent company Microsoft announced their joining of the MCP (Model Connection Protocol) Steering Committee. MCP is a standard proposed by the AI company Anthropic, aimed at better connecting AI models with data storage systems. This announcement marks the continued growth of MCP's influence in the AI industry; earlier, OpenAI and Google also expressed support for this standard.
Source attribution: Image generated by AI, licensed by Midjourney
The core function of MCP is to enable AI models to access information from various data sources to accomplish specific tasks. These data sources include business tools, software, content libraries, and application development environments. Through the MCP protocol, developers can build bidirectional connections between data sources and AI applications (such as chatbots). This means that developers can provide data to the outside world via "MCP servers" and create "MCP clients," such as applications and workflows, to connect with these servers when needed.
Microsoft and GitHub stated that they will provide comprehensive MCP support across their various platforms and services, including Microsoft Azure and Windows 11. It is expected that MCP integration will be achieved within the Windows system in the coming months, allowing developers to expose the functionality of their applications to AI models that support MCP. Microsoft noted in relevant materials that developers will be able to encapsulate required functions and capabilities within their applications as MCP servers to provide services, including Windows system features like file systems, window management, and Windows subsystems.
In terms of security, Microsoft stated that its identity and security team collaborated with Anthropic, the MCP Steering Committee, and the broader MCP community to design an updated authorization specification. This specification allows MCP-connected applications to safely access MCP servers, enabling users to grant permissions for AI applications to access data and services using "trusted login methods," including personal storage drives and subscription plans.
Meanwhile, GitHub also designed a registration service for MCP servers in collaboration with the MCP Steering Committee. This service allows developers to centrally store public or private MCP server entries, promoting the discovery and management of various MCP implementations and their configurations.
Key points:
🌐 GitHub and Microsoft join the MCP standard Steering Committee to promote the connection between AI models and data sources.
🔗 The MCP protocol enables developers to build bidirectional connections, supporting interaction between multiple data sources and AI applications.
🔒 Microsoft updates authorization specifications to ensure MCP connection security, while GitHub introduces a registration service for MCP servers.
