Teleport announces general availability of Secure Model Context Protocol (MCP)

Read full announcement here: https://goteleport.com/about/newsroom/press-releases/teleport-announces-general-availability-of-secure-model-context-protocol-mcp/?source=nhimg

With nearly half of enterprises (44%) already deploying AI, securing how AI agents interact with corporate data has become a top priority. Teleport has announced the general availability of Secure MCP, a new capability within the Teleport Infrastructure Identity Platform that applies enterprise-grade identity, access control, and governance to AI interactions with databases, MCP servers, and applications.

The Model Context Protocol (MCP)—an open standard likened to a “USB-C port for LLMs”—simplifies integration between AI models and tools. However, MCP lacks built-in access control, creating potential pathways for unauthorized data access. Teleport’s Secure MCP closes this gap by enforcing the same Zero Trust security guardrails used for human and non-human identities:

• Granular RBAC/ABAC to define exactly what AI agents can access

• Zero Trust Networking for authenticated, encrypted MCP connections

• Just-in-Time Authorization to prevent overprivileged access

• Comprehensive Audit Trails for compliance and forensic investigations

• Automated MCP Tool Discovery across hybrid environments

By extending its trusted computing architecture to LLM-driven workflows, Teleport enables enterprises to confidently adopt AI without sacrificing security or compliance.

Industry experts agree: as AI agents increasingly touch core systems, identity teams must prepare now. Teleport’s Secure MCP provides the policy enforcement, governance, and visibility needed to securely deploy AI agents at scale—aligning innovation with strong data protection.

Bottom Line

Teleport Secure MCP bridges the gap between AI innovation and enterprise security, giving organizations a unified way to control and audit AI-driven access alongside all other infrastructure identities.