Zenity Raises $125 Million to Secure 1 Billion AI Agents
Related Stories
Image credit : zenity
AI agents are moving from small workplace experiments into business systems where they can access data, use software, write code and take actions. That shift is creating a security problem that traditional chatbot controls were not designed to manage.
Zenity has raised a $125 million Series C led by Norwest. New investors Qumra Capital, SoftBank Vision Fund 2, Hitachi Ventures and LG Technology Ventures joined the round, alongside existing backers Vertex Ventures, Third Point Ventures, DTCP and Intel Capital. The funding will support product development and international expansion, particularly across Europe and Asia-Pacific.
Founded by Ben Kliger and Michael Bargury, Zenity provides an AI agent security and governance platform for large organizations. It helps security teams discover agents, understand their permissions and monitor what they are trying to do across enterprise systems.
The platform goes beyond reviewing prompts. Zenity says it can assess an agent’s intent and deterministically allow, change or block an action before it happens. This becomes important when an agent can access customer records, internal files, development environments or cloud infrastructure.
Zenity supports agents built across platforms including Microsoft Copilot, ChatGPT Enterprise, Gemini, Claude, Codex, Cursor, AWS Bedrock, Microsoft Foundry and Google Vertex AI. This gives companies one governance layer across commercial tools, coding agents and internally developed AI systems.
The company says most of its customers are Fortune 500, Global 2000 and other large organizations operating in regulated industries. Zenity also reports that revenue has tripled in each of the past two years and is on track to triple again in 2026. These figures are company-reported.
Its research division, Zenity Labs, has examined attacks in which malicious documents, calendar invitations or connected applications manipulate AI agents. Its AgentFlayer research demonstrated how zero-click attacks could hijack agents, alter workflows and expose enterprise data.