Google Announces Official Support for MCP Across Services and Cloud Products
Google announced official support for the Model Context Protocol (MCP) on Wednesday.
The update applies across its services and Google Cloud products.
The company says MCP will run on fully managed remote servers.
As a result, developers can connect AI agents more easily.
These agents include Gemini and other AI models.
The servers connect to real-world tools and data sources.
They also integrate with enterprise systems.
Therefore, developers and organisations can link AI agents to many third-party data sources.
Google says support will roll out to all services gradually.
Google Enables Managed MCP Servers
In a blog post, Google confirmed the first phase of the rollout.
Four services now support MCP servers.
These are Google Maps, BigQuery, Google Compute Engine (GCE), and Google Kubernetes Engine (GKE).
Thus, AI agents can call these platforms and use their data with permission.
This allows tasks to be completed using real-world information.
Google offered an example involving BigQuery.
A BigQuery MCP server can interpret table schemas.
It can run queries directly on enterprise data.
It can generate insights without moving data into an AI model’s internal memory.
Maps will also gain MCP functionality.
The server will provide location data such as weather, routes, and points of interest.
Consequently, AI agents can answer travel-related queries with accurate, updated information.
MCP Support Expands Beyond Google Services
Google is extending MCP support to Apigee.
Apigee manages APIs for many enterprises.
With this integration, existing APIs can become MCP-compatible.
Organisations will not need to rewrite or rebuild their workflows.
Agents can then use internal systems such as customer databases, workflow tools, and business logic.
Additionally, enterprise governance and security controls will remain intact.
On security, Google says multiple protections are in place.
Admins can manage access through Identity and Access Management (IAM).
They can track agent activity using audit logging.
They can also apply “Model Armor” to reduce risks such as indirect prompt injection.
What MCP Does
MCP is an open-standard protocol created by Anthropic.
It is described as a “USB-C port for AI.”
It defines how AI models connect to APIs, tools, and data systems.
The protocol uses a unified workflow for discovery, authentication, and access.
Before MCP, developers built custom connectors for each API.
This process was slow and fragile.
Now, AI clients such as Gemini CLI and AI Studio can call remote MCP servers in a consistent way.
Google Simplifies MCP Deployment
Previously, developers relied on community-built or open-source servers.
These required manual installation and ongoing maintenance.
With Google’s new managed servers, the company handles the infrastructure.
Developers simply connect their AI agents.
This improvement enables Gemini-powered agents to access enterprise-ready endpoints.
It also creates a globally consistent system across Google and Google Cloud services.