Zentera Systems Unveils Ensage™ AI, a Zero Trust Security Solution for the Age of Autonomous AI Agents

Zentera Systems, the pioneer in overlay Zero Trust security for enterprise and AI workloads, today announced Ensage AI, a comprehensive security platform that extends Zero Trust principles to discover, govern, and contain autonomous AI agents operating in enterprise networks.

AI coding agents such as Claude Code, Cursor, GitHub Copilot, and others are being adopted rapidly by engineering teams, and often in an ad-hoc fashion. These agents function autonomously, choosing what to share with LLMs, which tools to invoke, and which data to include in their requests, and perform those operations significantly faster than a human can oversee. Existing security tools were designed for human users and deterministic applications. They leave a critical governance gap: security teams cannot see which agents are running, what those agents are doing, or whether they are operating within policy.

Ensage AI closes this gap. It extends Zentera’s core CoIP® Overlay and Virtual Chamber technologies to provide a purpose-built governance framework for agentic AI that covers the full security lifecycle: visibility, identity, policy, enforcement, audit, and agent lifecycle management. This allows enterprises to apply Zero Trust principles and architectures to agentic AI.

“The challenge security teams face today is not whether AI agents are on their network. They already are,” said Jaushin Lee, President and CEO of Zentera Systems. “The challenge is that they have no way to distinguish a sanctioned agent performing an approved task from an unauthorized agent accessing sensitive systems. Ensage AI gives them that clarity – and the controls to act on it – all based on Zero Trust principles.”

Discover. Govern. Contain. Audit.

Ensage AI introduces four core capabilities:

Discover. A lightweight endpoint sensor detects AI agents, identifying each by its process signature, network behavior, and configuration. Security teams gain a real-time inventory of every AI agent in their environment, including which users are running them, which LLM endpoints they contact, and which tools and MCP servers they have installed.

Govern. A centralized policy engine lets security teams place agents inside a network sandbox, controlling what each agent is allowed to do and which services and endpoints it can reach. Integration with enterprise identity providers enables policy enforcement tied to existing users, groups, and projects. Centralized credential management replaces personal API keys with fully auditable enterprise credentials that never touch the endpoint.

Contain. Drawing on Zentera’s proven Virtual Chamber technology, Ensage AI creates per-agent network sandboxes that restrict each agent to only its authorized destinations. While Chambers protect trusted assets in untrusted environments, Sandboxes contain untrusted agents and protect the environment from them. If an agent violates policy or its approval is revoked, it can be quarantined instantly to limit unauthorized network access.

Audit. AI sessions are routed via the CoIP overlay network to the Ensage AI Session Controller (ASC), which examines every request and response between agents and external services. It enforces policies in real time and logs every interaction. All session data remains on premises within the customer’s environment. Configurable logging depth gives security teams the choice between simple metadata or a full audit trail for compliance and incident response.

Zentera Labs Intelligence

Ensage AI is powered by Zentera Labs Intelligence, Zentera’s dedicated agent security research operation. The Zentera Labs team continuously profiles the AI agent ecosystem – fingerprinting agents across platforms, tracking behavioral and protocol changes across releases, ensuring security keeps pace with an agent landscape that can change daily.

Designed for IP-Intensive Industries

Ensage AI addresses the governance, risk, and compliance requirements of industries where intellectual property protection is paramount. For example, in semiconductor and EDA environments, engineers routinely handle sensitive design data governed by export controls and customer NDAs. The platform enforces project-level isolation, ensuring that agents working on one program cannot access data or resources belonging to another.

“Other solutions focus on the tip of the iceberg: what the agent was prompted to do and what it outputs,” said Chase Cunningham, known in the cybersecurity industry as Dr. Zero Trust. “Ensage AI enforces Zero Trust under the surface – at the execution layer, where agents act, access, and cause harm.”

Availability

Ensage AI will be available in early access to select customers beginning in late March 2026. For more information, visit https://www.zentera.net/ensage-ai or contact sales@zentera.net.

About Zentera Systems, Inc.

Zentera Systems is the Zero Trust security company that protects what moves across enterprise networks – whether it is a user, a workload, or an AI agent. The company’s solutions deploy as an overlay on top of any infrastructure – IT, OT, cloud, or hybrid – to enforce Zero Trust security without re-architecting the network. With the launch of Ensage AI, Zentera now extends this same foundation to secure the autonomous AI agents operating inside enterprise environments. Zentera is headquartered in Silicon Valley and trusted by Global 2000 enterprises across semiconductor, financial services, healthcare, and public sector.