Versa Networks has launched a zero-trust framework embedded in its Model Context Protocol to rigorously verify each AI agent action before execution in network operations. Designed to close the trust gap in enterprise AI systems, this new architecture enforces strict identity, role-based, and policy-driven controls, and mandates human approval where required, ensuring full transparency and compliance.
- Zero-trust MCP validates every AI agent action against identity, roles, and policies
- Human approval required for high-risk or policy-restricted operations
- Fully integrated with VersaONE platform and AI co-pilot Versa Verbo
What happened
Versa Networks announced a new zero-trust architecture for its Model Context Protocol (MCP) that verifies all AI agent actions before they execute within network operations. This architecture is part of Versa Verbo, the company’s AI-powered operations co-pilot, and is fully integrated into the VersaONE Universal SASE Platform. The system is designed to prevent unauthorized or opaque AI-driven changes by enforcing identity verification, role-based access control, and pre-defined policy checks for every action.
Why it matters
As enterprises increasingly deploy AI agents to automate network and security management, concerns grow over the lack of visibility and control over these nonhuman identities. Traditional secure access service edge (SASE) and security service edge (SSE) platforms were not originally designed to handle AI agents’ unique risks and operational profiles. Versa’s zero-trust MCP architecture addresses this challenge by treating AI agent actions with the same scrutiny as human users’ activities, minimizing operational risk and helping maintain compliance.
This approach aligns with Gartner analysts’ warnings that AI introduces a high-volume class of digital users lacking in traditional security frameworks. By ensuring every AI action is verified, logged, and, where necessary, human-approved, Versa helps enterprises confidently adopt AI-driven network operations while mitigating potential liabilities associated with unmonitored agentic AI activities.
What to watch next
Versa plans to evolve this zero-trust architecture by moving beyond per-action validation toward more comprehensive policy-driven automation, adapting to broader enterprise adoption of AI operations. As they advance, enterprises can expect tighter integration of AI event correlation, anomaly detection, and guided troubleshooting to streamline AI agent workflows without sacrificing security or auditability.
This development builds on Versa’s previous open-source MCP Server release and incremental agentic capabilities added to Versa Verbo last year. As the company continues expanding its AI-driven network operation tools, industry observers should watch how zero-trust controls for AI agents become a fundamental requirement for securing automated enterprise environments.