Versa Networks Unveils Zero-Trust Architecture for AI Agents

In a significant advancement for enterprise AI security, Versa Networks Inc. has introduced a zero-trust architecture for its Model Context Protocol (MCP). This new framework validates every action taken by AI agents in its operations co-pilot, Verbo, before those actions are executed, effectively addressing what the company identifies as a 'trust gap' in agentic AI.

The rise of AI agents in enterprise settings has prompted a reevaluation of existing security protocols. Traditional secure access service edge (SASE) platforms, as highlighted in a recent report by Gartner Inc., are not equipped to handle the new, high-volume class of digital users represented by AI agents. These platforms were not designed for the complexities of nonhuman identities, making the need for innovative solutions critical.

Under Versa's new architecture, no action generated by an AI agent is considered implicitly trustworthy. Each action undergoes rigorous checks against user identity, role-based access controls, and predefined system policies. Human approval is required for actions deemed sensitive or risky, ensuring that administrators maintain oversight. This level of scrutiny not only enhances security but also provides full attribution for every approved action, which is essential for compliance and auditing purposes.

Sridhar Iyer, senior director of AI and machine learning at Versa, emphasized the urgency of this development, stating, "Enterprise AI is at an inflection point. Until every agent action can be verified, AI in production can turn into a liability, not an advantage. Versa is solving that head-on and we believe zero trust is the only viable architecture for trustworthy agentic AI."

The introduction of this zero-trust execution layer aligns with Versa's ongoing commitment to enhancing agentic operations. Earlier iterations of the MCP Server, launched in April 2025, provided secure access to Versa's tools and APIs. The current advancements in Verbo mark a progression toward more comprehensive policy-driven automation as enterprise adoption of AI technologies matures.

Verbo now combines the new zero-trust architecture with features such as AI-driven event correlation, anomaly detection, and guided troubleshooting. This integration aims to allow enterprises to run AI-driven operations without sacrificing the oversight required by compliance and security teams.

Versa has proactively secured substantial funding to support its growth, amassing $316 million from a variety of investors, including Sequoia Capital and BlackRock. This financial backing underscores the market's confidence in Versa's potential to reshape AI operations in enterprise environments.

With the Zero Trust MCP architecture now available as part of the VersaONE Universal SASE Platform Release 23.1.1, enterprises can implement these advanced security measures immediately, positioning themselves to better navigate the complexities introduced by AI integration in their operations.

As enterprise AI continues to evolve, the introduction of zero-trust frameworks like Versa's could set a new standard for security in AI-driven operations, ensuring that trust is built into the very fabric of digital interactions.