This is the first post in a series kicking off our journey to deliver Model Context Protocol (MCP) server support. MCP is evolving quickly. We will share updates as the ecosystem matures and as our products grow alongside it.
Leaders in AI know that interoperability is key to unlocking agentic innovation. Until now, the path has often been unclear, with limited natural language support and fragmented point-to-point integrations. The vision is simple: employees and AI applications working together through natural language interfaces, while maintaining security, governance, and auditability. Today, we are excited to introduce the ServiceNow MCP Platform Manager, including the ServiceNow QuickStart MCP Server, to begin delivering on that vision; it's live now through our GenAI Innovation Lab!
What is MCP?
The Model Context Protocol (MCP) is an open standard for connecting AI applications to external systems. MCP servers expose functionality as structured tools, resources, and prompts. Tools are particularly valuable because they let AI applications perform specific functions through structured requests and responses. This provides a reliable way for AI applications to act consistently across systems without requiring one-off integrations.
MCP Platform Manager
MCP Platform Manager is the administrative experience to publish, secure, and observe MCP tools and servers. You decide which ServiceNow capabilities are surfaced, how they are surfaced, and who can access them. Flexibility and admin control are key differentiator in the market.
QuickStart MCP Server
The QuickStart MCP Server is a fully functional MCP server for ServiceNow with ready-to-use tools, giving you a fast path to testing and learning.
Together, these form a governed interoperability layer that allows MCP clients and AI agents to act safely across ServiceNow and connected systems under your control through the protocol.
We believe that MCP server governance should be easy for AI admins everywhere.
Powerful tools for agentic interoperability
With MCP, trusted ServiceNow capabilities can be packaged into tools that AI applications call directly.
Our currently live Gen AI Innovation Lab places Now Assist Skills at the center, making them the first set of ServiceNow capabilities available as tools through MCP Servers. Skills are an ideal starting point because they are designed for AI use, operate within clear governance boundaries, and can be customized as needed. Our targeted December GA will follow this same approach.
Create bespoke servers with Skills as Tools.
Looking ahead, we plan to make Build Agent callable through MCP. This will allow developers to scaffold apps, debug, and extend solutions from the tools they already use. By pairing Build Agent with MCP Server, customers gain orchestration power and a new way to innovate on the ServiceNow AI Platform.
Beyond Build Agent, we are evaluating additional ServiceNow capabilities and constructs for the future, such as Flows, Knowledge Graph, and Scripted REST APIs. We strongly welcome input on which tool types are critical for your use cases and matter most to you.
The power of Now in any employee experience
The MCP Server extends ServiceNow’s trusted functionality into tools that can be invoked externally, enriching any employee experience wherever it is needed.
For example, it can generate governed summaries of incidents to streamline handoffs and provide clear status updates, or automatically create knowledge articles from resolved and closed cases so insights are captured and reused.
In both scenarios, the power of Now extends beyond a single UI while still being governed from a central control point. Many enterprises are beginning with lower-risk patterns like these to validate value, governance, and security before expanding into more advanced automation.
Trust when it matters most
The defining advantage of the ServiceNow MCP Server is trust. Unlike third-party MCP servers, only ServiceNow provides enterprise-grade governance, maintainability, and observability:
- Scoped exposure of tools by server and catalog, so only the right clients see the right tools
- Identity and access through OAuth and ServiceNow controls, so you know who is making each call
- Observability across usage, reliability, and latency by tool and by server
- Cost transparency, including Assist consumption
This work aligns with ServiceNow AI Control Tower, so MCP activity can be observable and governable in the broader AI program. We will share more as we continue to integrate telemetry and policies across the stack.
Why it matters
The ServiceNow MCP Platform Manager and ServiceNow QuickStart MCP Server are more than new features. They are the foundation for safe, enterprise-ready agentic interoperability. They bring together three essential capabilities:
- Powerful tools that connect AI agents and workflows.
- The power of ServiceNow available in any employee experience.
- Built-in guardrails that ensure enterprises stay secure, auditable, and in control.
The ServiceNow MCP Platform Manager is the governed tool plane built for the AI era.
