Here’s the translation to American English:
In the context of Microsoft Build 2025, the technology company has revealed its latest innovation: the integration of the Model Context Protocol (MCP) into Windows 11. This initiative aims to establish a secure and standardized communication framework between tools and artificial intelligence (AI) agents, addressing the growing concerns about security in digital interactions.
The MCP is described as a lightweight and open protocol that enables applications and agents to discover and invoke tools uniformly. This feature facilitates orchestration between services, both local and remote, allowing developers to build solutions that can integrate across various platforms. Within this new ecosystem, the protocol classifies tools into three fundamental roles: hosts, which are applications like Visual Studio Code; clients, which initiate requests; and servers, which provide specific functionalities.
With the adoption of the MCP, Windows 11 promises to empower developers to create intelligent applications that leverage generative AI capabilities. This will allow tools to act on behalf of the user, simplifying processes and improving efficiency. However, this openness also presents a series of security risks that must be managed carefully. MCP servers, if not configured properly, can be vulnerable to unauthorized access or command injection attacks, which expose sensitive data.
Microsoft has identified several emerging threat vectors, such as malicious content injection and credential leakage. This underscores the need to implement robust security controls. In response to these concerns, Windows 11 will introduce measures such as communication mediated by a trusted proxy, which centralizes authentication, and tool-level authorizations, requiring explicit user approval for each interaction.
Additionally, MCP servers will need to meet essential security requirements, including code signing and security testing of exposed interfaces, to be accepted into the central registry of Windows 11. Microsoft plans to share a private preview of these capabilities with developers, ensuring that security evolves alongside the implementation of these innovations.
As it advances in the development of its platform, Microsoft reaffirms that trust is a key pillar of innovation. By integrating security measures at the core of its agent strategy, the company aims to ensure that the future of artificial intelligence on Windows is not only advanced but also secure.
via: MiMub in Spanish