Noma Expands Security Features for AI Agents on Employee Workstations

Noma Enhances Endpoint Security for AI Agents

In an age where enterprise environments are increasingly beset by cybersecurity threats, Noma is taking significant steps to elevate security across its platform. On September 28, 2026, the company unveiled its latest update, designed to broaden the security measures applied to AI agents operating on employee endpoints. This extension is critical, given that the endpoint has become one of the largest blind spots in enterprise AI security.

Noma is recognized as a leader in the field of AI and agent security. Its platform has observed that agents, MCP (Multi-Cloud Platform) servers, and various AI skills already in operation on employee devices are often given the same permissions and credentials as the personnel who installed them. This poses considerable risks, especially when malicious actors gain access to these credentials. Noma's enhanced capabilities will allow organizations to discover and monitor these agents effectively, enforce access control measures, and utilize AI for real-time threat detection and response (AI-DRAI).

Expanded Capabilities

The new features from Noma enable security teams to maintain oversight of agents, skills, and MCP servers running on employee machines, controlling their activities seamlessly across various environments, whether they be SaaS, on-premise, or hybrid models. These capabilities are a game-changer for security protocols, as they provide a unified approach to managing the agent landscape, allowing organizations to have a single control plane across all agent assets.

Developers often utilize coding agents like Claude Code and Cursor to integrate with source code and production systems, while business personnel rely on productivity agents such as Claude Cowork. With this updated technology, organizations can readily identify which agents are operational, assess their permissions, and set comprehensive policies governing their operations.

Risk Mitigation and Runtime Protection

The update does not merely improve discovery and governance; it also incorporates rigorous runtime protection measures. The AI-DR component continuously monitors agent behavior, establishing behavioral baselines while detecting potential anomalies. Whether it's prompt injections, data leaks, or unauthorized access, the system is designed to catch threats before they escalate into full-blown security incidents. Moreover, the contextual policies provided enable organizations to track actions across sessions, which helps identify threats that may culminate from a series of approved actions turning into malicious tasks.

Noma co-founder and CEO Niv Braun emphasized the importance of this new implementation by stating, "Agents on the endpoint carry some of the most privileged identities in the company, often unnoticed by security until something goes wrong. By extending our access control and AI-DR model to the endpoint, security teams gain the same visibility and enforcement as they have across SaaS and homegrown agents."

Open Enforcement for Flexible Coverage

The updated model leverages Open Enforcement to apply policies consistently across various platforms already in place within organizations, including codings hooks and AI gateways. With the inclusion of hundreds of pre-tuned AI-DR policies reflective of real-world operations of Fortune 500 security teams, Noma is equipping companies with the tools they need to safeguard their sensitive data effectively.

Noma's approach to agent security extends to various popular agents such as Claude Code and Codex, ensuring that organizations can establish a robust, adaptive security framework tailored to their specific needs. As employee devices become hotspots of potential security vulnerabilities, these enhancements promise to secure the endpoints effectively, allowing organizations to operate with greater confidence in their security posture.

For more information on how Noma is enhancing endpoint security for agents, you can visit their website here.

About Noma

Noma is dedicated to delivering exceptional AI and agent security solutions that protect enterprises across diverse environments, from cloud-based platforms to localized infrastructures. Partnered with notable venture capital firms and widely trusted by Fortune 500 organizations, Noma continues to set industry standards for AI trust, risk, and security management (AI TRiSM). Learn more at Noma's website.