CIS Guide Recommends Controls to Secure Model Context Protocol
Summary
CIS, Astrix Security, and Cequence Security developed AI security companion guides. The MCP guide recommends enforcing authorization outside the model, securing transports, governing capability changes, and validating tools and inputs.
Key points
- The MCP Companion Guide focuses on the protocol layer where AI systems access data, invoke tools, and execute requests.
- Servers or gateways—not models—should enforce identity, token validation, access controls, and least privilege.
- Recommended transport protections include restricting bind addresses, using TLS, validating HTTP Origin headers, and preventing DNS rebinding.
- Organizations should track MCP capability changes with allowlists, signed artifacts, baselines, and deny-by-default policies.
- The guide recommends schema validation, parameter sanitization, and review and versioning of tools, resources, and prompts.
- CIS developed the MCP guide alongside LLM and AI Agent companion guides to extend CIS Controls to AI environments.
Article Details
- Event Type
- Publication of cybersecurity guidance for securing the Model Context Protocol (MCP) and AI environments.
- Impact
- The guidance recommends controls for MCP authorization, transports and endpoints, capability changes, and tool, resource, and prompt hardening to reduce risks including unauthorized access, destructive operations, prompt injection, and data leakage.
Vendors
Astrix SecurityAs a result, the Center for Internet Security® (CIS®), Astrix Security, and Cequence Security partnered to develop actionable cybersecurity guidance tailored to AI environments.Center for Internet SecurityAs a result, the Center for Internet Security® (CIS®), Astrix Security, and Cequence Security partnered to develop actionable cybersecurity guidance tailored to AI environments.Cequence SecurityAs a result, the Center for Internet Security® (CIS®), Astrix Security, and Cequence Security partnered to develop actionable cybersecurity guidance tailored to AI environments.
Products
AI Agent Companion GuideThe result is three CIS Companion Guides: the AI Large Language Model (LLM) Companion Guide, the AI Agent Companion Guide, and the Model Context Protocol (MCP) Companion Guide.AI Large Language Model (LLM) Companion GuideThe result is three CIS Companion Guides: the AI Large Language Model (LLM) Companion Guide, the AI Agent Companion Guide, and the Model Context Protocol (MCP) Companion Guide.CIS Critical Security ControlsThe work extends the globally recognized CIS Critical Security Controls® (CIS Controls®) into environments where autonomous decision‑making, tool and API access, and automated threats introduce new risks.Model Context ProtocolThe Model Context Protocol (MCP) provides a structured, open interface through which models and agents invoke tools, access data sources, and apply reusable prompts using a consistent, standardized mechanism.Model Context Protocol (MCP) Companion GuideThe result is three CIS Companion Guides: the AI Large Language Model (LLM) Companion Guide, the AI Agent Companion Guide, and the Model Context Protocol (MCP) Companion Guide.