Zero Trust Security for AI Agents and Multi-Party Computation Workflows
Zero Trust Principles Extended to AI Agents and MCP Workflows
In a significant development, Versa has unveiled a pioneering zero-trust architecture for the Model Context Protocol (MCP), effectively integrating zero-trust principles into AI execution. This groundbreaking approach ensures that every AI-generated action is meticulously validated against user identity, role-based access controls, and system policies before execution.
Addressing Challenges Associated with Agentic AI Systems
As enterprises increasingly adopt agentic AI systems, the challenges associated with maintaining visibility and control have become more pronounced. A single prompt can inadvertently trigger multiple actions across network and security environments, creating blind spots and making it difficult to monitor tasks accurately. Moreover, systems can misinterpret intentions or generate unintended actions, further exacerbating these issues.
“AI has introduced a new, high-volume class of digital users in the form of agents that traditional SSE/SASE Platforms were not built to secure.” — According to a recent report by Gartner.
Versa’s Zero-Trust MCP Architecture Implementation
In response to this pressing concern, Versa’s zero-trust MCP architecture has been implemented within Versa Verbo, the company’s AI-powered operations co-pilot. This innovative solution is seamlessly integrated with the VersaONE Universal SASE Platform, offering comprehensive coverage and protection.
Administrators define policies in advance that dictate which agent actions execute automatically, which require human approval, and which are blocked, taking into account user identity, role, system context, action type, and risk level. Each approved action is meticulously logged with full attribution, granting operators unparalleled visibility, auditability, and policy-driven control without hindering AI-driven operations.
"We’re at a pivotal moment in the evolution of enterprise AI," remarked Sridhar Iyer, Senior Director of AI and ML at Versa. "Until every agent action can be verified, AI in production can become a liability rather than an asset. Versa is proactively addressing this issue through our commitment to zero-trust principles, and we firmly believe this is the only viable architecture for trustworthy agentic AI."
Milestone in Versa’s Multi-Year AI Innovation Strategy
This announcement marks a significant milestone in Versa’s multi-year AI innovation strategy, building upon the successful launch of its open-source MCP Server in April 2025. This pioneering initiative enables secure access between Large Language Models (LLMs) and Versa tools and APIs, enhancing contextual awareness and operational efficiency. The subsequent introduction of agentic capabilities in Verbo in late 2025 has further solidified Versa’s position as a leader in the field.
Moving forward, Versa will continue to push the boundaries of AI innovation, focusing on extending its capabilities from per-action validation towards policy-driven automation and deeper operational visibility as enterprise adoption of AI matures. With Versa Verbo supporting SaaS, on-premises, and hybrid deployments, organizations now have unprecedented control over model selection, data locality, and processing environments, ensuring a seamless and secure experience for all stakeholders involved.