EVERYTHING
NetFoundry announced the expansion of its AI Enclave solution with the addition of zero trust MCP and LLM gateways. These products are engineered to enable secure, scalable deployment of agentic AI across various environments, including on-premise (even air gapped), hybrid, and cloud infrastructures.
According to Prnewswire, these new Gateways leverage NetFoundry's Identity-First Reachability™ framework. This approach grants every AI agent, MCP server, and LLM endpoint a sovereign machine identity. Crucially, while agents are authenticated, they are never given direct access to API keys, service accounts, or shared secrets, effectively eliminating the traditional attack surface.
Addressing the Core Security Problem of AI Reachability
The launch directly confronts what NetFoundry identifies as the fundamental security vulnerability of the AI era: reachability. Every exposed endpoint—be it an API, an LLM, or an MCP server—represents a potential entry point for attackers. Galeal Zino, CEO and Founder of NetFoundry, stated that "Vulnerability exploitation is the #1 breach vector today because attackers go after what they can reach."
The proliferation of AI agents, coupled with broadly reachable MCP servers and the rapid distribution of secrets across teams, significantly expands this attack surface. The Gateways counteract this by making the entire AI infrastructure invisible by default while blocking lateral movement.
- Identity-First Security: Authorization occurs at the service level, not merely the network level.
- Secure Communication: All connections are initiated outbound and remain end-to-end encrypted.
- Operational Efficiency: Organizations can save up to 50% in AI token costs while maintaining robust governance.
The Limitations of Existing Security Architectures
This shift toward identity-centric security is particularly relevant given the challenges faced by conventional enterprise architectures. Gartner highlighted that securing AI applications remains a nascent market, noting that many existing SASE (Secure Access Service Edge) architectures struggle with applicability. This difficulty arises because servers are often not directly connected to the SASE fabric and AI applications frequently use direct API calls that bypass traditional control points.
The new MCP and LLM Gateways deliver specific capabilities designed for this complex environment: they provide identity, visibility, control, and governance across distributed systems. Kevin Day, CTO of Rhapsody, noted how NetFoundry provides the secure network foundation needed to support private, policy-based access across diverse environments, including emerging AI-enabled workflows.
By ensuring that unauthorized agents or cyber attackers cannot reach the MCP and LLM gateways, NetFoundry establishes a secure perimeter around dynamic AI interactions. This allows enterprises to deploy at software speed without simultaneously handing over a larger target to malicious actors, providing a complete security layer for modern digital transformation.