Versa's introduction of a zero trust architecture for the Model Context Protocol (MCP) marks a significant step in securing AI execution. This patent-pending approach validates every AI-generated action against user identity, role-based access controls, and system policies before implementation. Human approval is required for actions specified by administrators, addressing the risks associated with deploying agentic AI.
The need for such a framework has become clear as enterprises deal with the complexities of AI systems. A single prompt can trigger a series of actions within network and security environments, often obscuring visibility into the processes involved. Misinterpretation of user intent can lead to unintended outcomes, raising operational and security issues that organizations cannot ignore. Gartner highlights that traditional security platforms were not built to handle the new high-volume class of digital users introduced by AI agents.
Versa's zero trust MCP architecture integrates into the company's AI-powered operations co-pilot, Versa Verbo, and is part of the broader VersaONE Universal SASE Platform. This combination equips enterprises with AI-powered event correlation and anomaly detection capabilities, allowing for a more controlled adoption of AI-driven operations in production settings. The architecture enables administrators to clearly define policies, specifying which agent actions can execute automatically, which require human oversight, and which should be blocked entirely.
Each approved action is logged meticulously, ensuring full attribution while providing operators with the visibility and auditability needed for effective policy-driven control. “Enterprise AI is at an inflection point,” said Sridhar Iyer, Senior Director of AI and ML at Versa. “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.”
This announcement continues Versa's multi-year strategy to enhance AI innovation, which began with the launch of its open-source MCP Server in April 2025. This initiative aimed to facilitate secure access between large language models and Versa's tools and APIs, improving contextual awareness and operational efficiency. The introduction of agentic capabilities in Verbo followed this foundational step, further evolving the platform's capabilities.
As enterprises increasingly adopt AI technologies, Versa plans to evolve its offerings beyond simple action validation to include policy-driven automation and a deeper understanding of operations. Versa Verbo supports various deployment models, including SaaS, on-premises, and hybrid solutions, giving organizations control over model selection, data locality, and processing environments. This flexibility is crucial as the demand for secure and efficient AI operations continues to grow.
Quick answers
What is the zero trust architecture introduced by Versa?
Versa's zero trust architecture for the MCP validates every AI action based on user identity and system policies, requiring human approval where necessary.
Why is zero trust important for AI-driven operations?
Zero trust is crucial as it ensures that every action by AI agents is verified to prevent operational liabilities and maintain control.
How does Versa’s solution enhance security for enterprises?
The solution enables clear policy definitions for AI actions, logs all approved actions for accountability, and integrates anomaly detection for improved oversight.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
