As organizations increasingly integrate AI into their operations, the challenge of trust and security in AI execution has become paramount. Enterprises face a dilemma: how to harness the productivity of AI agents while ensuring their actions are secure and verifiable. Versa has stepped in to address this issue with a patent-pending Zero Trust architecture specifically designed for the Model Context Protocol (MCP).
Versa's approach guarantees that no action taken by an AI agent is implicitly trusted. Each step initiated by these agents is validated against user identity, role-based access controls, and established system policies before execution. Human validation is mandated based on administrator-defined policies, adding a crucial layer of oversight. This model aims to eliminate risks linked to unintended actions or misinterpretations of intent that can arise in traditional AI operations.
The urgency of this innovation is highlighted by a recent Gartner report, which points out that AI has introduced a new class of digital users—agents that existing security platforms were not designed to protect. With AI systems often triggering multiple actions across critical networks, the absence of visibility and control can expose enterprises to significant vulnerabilities.
Integration with Versa's Offerings
The Zero Trust MCP architecture has been integrated into Versa Verbo, the company’s AI-powered operations co-pilot, and is complemented by the VersaONE Universal SASE Platform. This integration enables enterprises to implement AI-driven operations in their production environments without compromising security. Features such as anomaly detection, guided troubleshooting, and AI-powered event correlation allow organizations to maintain control over their AI applications.
Administrators can establish predefined policies that determine which agent actions can execute automatically, which require human approval, and which should be blocked altogether. This level of granularity ensures that every approved action is logged with full attribution, enhancing visibility, auditability, and control over AI operations. The architecture promotes operational efficiency without creating bottlenecks, a vital consideration in fast-paced environments.
A Shift in Enterprise AI Strategy
Sridhar Iyer, Senior Director of AI and ML at Versa, stated, “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.” This sentiment captures the essence of the current environment, where balancing innovation and security is critical.
Versa’s commitment to advancing its AI capabilities is evident in its multi-year innovation strategy, which began with the launch of the open-source MCP Server in April 2025. This initiative has established the groundwork for secure interactions between large language models (LLMs) and Versa's tools, improving contextual awareness and operational efficiency. With the introduction of agentic capabilities in Verbo in late 2025, Versa continues to focus on policy-driven automation and enhanced operational visibility as AI adoption grows across enterprises.
Availability and Future Outlook
The Zero Trust MCP architecture is currently available within the VersaONE Universal SASE Platform Release 23.1.1. This version supports various deployment models, including SaaS, on-premises, and hybrid solutions, providing organizations with the flexibility to manage their model selection, data locality, and processing environments.
As organizations navigate the complexities of AI integration, Versa’s latest offerings mark a significant advancement in ensuring that AI can be a tool for productivity without sacrificing security. The emphasis on Zero Trust principles in AI execution could redefine how enterprises approach their AI strategies, paving the way for safer and more reliable AI-powered operations in the future.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
