One of the key developments in the AI field in 2025 was the widespread adoption of the Model Context Protocol (MCP). Developed and launched by Anthropic in 2024, MCP has become an essential tool for enabling agents to interact securely and efficiently with external data and systems.
Now, Google has announced the launch of fully managed remote MCP servers that will provide a “unified layer across all Google and Google Cloud services.”
AI Agent & Copilot Summit is an AI-first event to define opportunities, impact, and outcomes with Microsoft Copilot and agents. Building on its 2025 success, the 2026 event takes place March 17-19 in San Diego. Get more details.
MCP Meets Google Services
Google has improved its existing API infrastructure to support MCP. This means that developers only need to direct their AI agents or MCP clients to a single endpoint for both Google and Google Cloud services. Additionally, the company has utilized Apigee, Google Cloud’s enterprise-grade API management solution, to allow developers to provide this functionality for purpose-built APIs designed for specific data flows and business logic.
In practice, this means that both purpose-built APIs and third-party APIs can be considered agent-discoverable sources. “Google’s support for MCP across such a diverse range of products, combined with their close collaboration on the specification, will help more developers build agentic AI applications,” said David Soria Parra, co-creator of MCP & Member of Technical Staff, Anthropic.
“As adoption grows among leading platforms, it brings us closer to agentic AI that works seamlessly across the tools and services people already use.”
MCP support for all Google services will be released incrementally. Here are the first to benefit:
Google Maps: Maps Grounding Lite leverages the MCP server to connect AI agents with geospatial data from Google Maps. This includes all the familiar capabilities, such as weather information, travel routing data, and location-specific knowledge. Companies can deploy AI agents that can access this information in real time and answer queries using the power of the Google Maps platform.
BigQuery: Google’s BigQuery MCP server allows AI agents to interpret schemas and execute queries against enterprise data assets natively, avoiding the potential efficiency loss and security risks of moving data elsewhere.
Google Compute Engine (GCE): MCP also supports the autonomization of infrastructure management, giving agents the power to manage infrastructure workflows from initial builds to dynamic adaptation.
Google Kubernetes Engine (GKE): The GKE MCP server provides a structured interface that AI agents can leverage to interact with GKE and Kubernetes APIs. Agents can operate either autonomously or with human involvement to analyze problems, perform remediation tasks, and optimize expenses.
Closing Thoughts
“Google is committed to leading the AI revolution not just by building the best models, but also by building the best ecosystem for those models and agents to thrive,” reads a Google press release. And this is an essential takeaway from the news around the company’s MCP integration.
MCP is a creation of Anthropic. However, Google, like other companies in the Cloud Wars Top 10, is actively embracing this technology to accelerate not only its own products but the entire agentic AI industry. The pace of advancements in the AI space has been so rapid that it’s often hard to remember where one technology started and how it evolved from its predecessor. Nonetheless, agentic AI is still in its early stages.
Connectivity is a critical part of the burgeoning AI agent ecosystem, and Google has made great strides with these integrations. They are not only starting with some of the company’s most popular services but also on familiar products, especially Google Maps. These tools will not only leverage MCP effectively but also demonstrate how MCP can be used to bring in data from familiar sources. At this fledgling stage, understanding agentic AI is as important as adopting it.
Ask Cloud Wars AI Agent about this analysis
