Onyx Security has integrated its secure AI control plane with Glean’s enterprise AI platform, giving organizations a way to inventory and govern agents built and operated on Glean. The companies say the integration is designed to extend security controls across agents as they are deployed and used across business departments.
Glean supplies the environment for building, deploying and operating agents, while Onyx provides governance and security capabilities. The integration centers on discovering Glean agents and recording their access and capabilities. Organizations can assign each agent an owner, business purpose and risk classification, according to the companies.
Inventory and policy oversight
Onyx says the integration can apply enterprise AI governance and compliance policies to agents. It also continuously assesses changes to the permissions, tools and data access available to each agent. That ongoing assessment is intended to keep governance tied to changes in how agents are configured, rather than limited to the initial deployment.
The companies describe the arrangement as a single governance model spanning deployment, ongoing operation and expansion. Onyx also maintains centralized audit evidence across AI environments, providing a record for oversight and compliance work. The announcement does not detail specific policy standards, integrations beyond Glean, or how audit records are stored.
Controls during agent operation
In addition to discovery and policy oversight, Onyx applies what it calls intent-aware runtime controls to agent prompts, responses and actions. The company says these controls can block unsafe requests. In other cases, they can steer an agent away from an unintended action and back toward company policy.
The approach addresses agent activity while the system is running, rather than focusing only on cataloging permissions in advance. Onyx co-founder and chief AI officer Gil Elbaz said security teams need visibility into what agents can do as agents gain access and autonomy. He said the integration combines Glean’s enterprise context with Onyx’s security control plane to govern agent activity across an organization’s AI environment.
Glean vice president of partnerships Zubin Irani described context and governance as foundational to making enterprise AI useful and trustworthy. He said the Onyx integration builds on that foundation as organizations deploy agents across their businesses. The announcement did not provide customer names, deployment numbers or performance results.
Company context
Onyx is an AI agent security startup. It said it had raised $153 million in total, including a $113 million Series B led by Bessemer Venture Partners in July 2026. The company launched in March 2026 with $40 million in seed and Series A funding.
Glean, an enterprise AI platform company, last raised $150 million in a Series F round in June 2025, at a reported valuation of $7.2 billion. The partnership brings together Glean’s agent-building and operating platform with Onyx’s stated focus on security controls and governance. The companies’ announcement positions the integration for organizations expanding autonomous, connected agents, while emphasizing inventory, permissions monitoring, runtime intervention and audit evidence as components of oversight.
