Google Announces Official Support for MCP Across Services and Cloud Products
Google announced full support for the Model Context Protocol (MCP).
The news was released on Wednesday.
The company said MCP will cover its services and Google Cloud products.
Short sentences, clear, in news style.
Google stated that MCP servers are fully managed.
Therefore, developers will find it easier to connect AI agents.
These agents include Gemini and other AI models.
Additionally, MCP can access real-world tools, data sources, and enterprise systems.
As a result, enterprises and developers can connect to more third-party data sources.
The company added that support will roll out gradually across all services.
Google Launches Managed MCP Servers
In a blog post, Google said the initial rollout covers four services.
These services include Google Maps.
They also include BigQuery, Google Compute Engine (GCE), and Google Kubernetes Engine (GKE).
Therefore, AI agents can now call these platforms.
They can access data (with permissions).
This allows them to perform tasks in real-world scenarios.
Google provided an example using BigQuery’s MCP server.
Agents can read table schemas.
They can run queries directly on enterprise data.
They can generate insights without moving data into the AI’s internal memory.
Similarly, the Maps MCP server provides real-world location data.
This includes weather, routes, and points of interest.
Therefore, agents can answer travel-planning questions.
The data is more reliable and more up-to-date.
Google Extends MCP Support to Apigee
Google is not limiting MCP to its own services.
The company is extending support to Apigee.
Apigee is a platform enterprises use to manage APIs.
Therefore, companies can convert existing APIs into MCP tools.
They do not need to rewrite or rebuild their systems.
As a result, AI agents can use internal databases, workflow systems, and business logic.
Meanwhile, enterprise governance and security policies remain in place.
On security, Google said multiple protections are in place.
Admins can control access using IAM.
They can use audit logs to track agent actions.
They can also apply “Model Armor” to mitigate threats such as indirect prompt injection.
What Is MCP?
MCP was developed by Anthropic.
It is an open-standard protocol.
It is often compared to a “USB-C port for AI.”
It defines how AI models connect to external data, APIs, tools, and services.
It uses a unified format and workflow.
In the past, developers built custom connectors for every API.
The task was slow and fragile.
With MCP, clients such as Gemini CLI and AI Studio can call remote servers.
They can discover, authenticate, and use external resources in a standard way.
Previously, Google’s MCP support relied on community-built servers.
Developers had to install and manage them themselves.
Now, Google provides managed remote MCP servers.
The company handles all the infrastructure.
Developers simply connect.
This allows Gemini-powered agents to use globally consistent, enterprise-ready Google service endpoints.
Leave a Reply