Zenity raises $125 million to secure autonomous AI agents

Why it matters

The round funds Zenity's bid to become the independent control layer for agents that can access data and execute work across enterprise systems. Its central risk is platform consolidation: cloud and security incumbents can build similar controls into products customers already buy.

Ben Kliger (@benkliger) and Michael Bargury (@mbrg0) have raised $125 million for Zenity, financing an expansion of the security platform they began building before autonomous AI agents became an enterprise spending category.

Zenity announced the Series C in a post on X on August 3rd. Norwest led the round, with new investors Qumra Capital, SoftBank Vision Fund 2, Hitachi Ventures and LG Technology Ventures. Existing backers Vertex Ventures, Third Point Ventures, DTCP and Intel Capital also participated, according to Zenity's announcement.

The financing follows Zenity's $38 million Series B in October 2024, co-led by Third Point Ventures and DTCP. Zenity said that earlier round brought its total capital raised above $55 million, meaning the Series C pushes its disclosed funding beyond $180 million. Zenity did not state a valuation or identify other financial terms.

The round gives Kliger and Bargury a large balance sheet for a security problem that is moving from experimental AI deployments into operational infrastructure. Enterprises are connecting agents to internal data, cloud services, software-development environments and business applications. Those connections allow agents to execute tasks, invoke tools and move information across systems, creating risks that conventional identity, application and data-security products were designed to observe only in pieces.

A thesis formed inside Microsoft

Kliger and Bargury met while working in Microsoft's cloud security division. Kliger spent nearly six years at Microsoft after roles in Israel's Unit 8200, the Israeli Prime Minister's Office, Deloitte and Fortscale, which RSA acquired. Bargury worked as a senior architect in Microsoft's Cloud Security CTO Office and led security efforts spanning APIs, internet-connected devices, infrastructure-as-code and confidential computing.

Kliger has said the pair watched employees create applications, automations and bots outside the usual software-development process. Business users could connect sensitive information and production systems with limited visibility from security departments. Zenity emerged in 2021 around that governance gap, initially concentrating on low-code and no-code software.

The arrival of enterprise copilots and autonomous agents enlarged the same problem. Software creation was spreading beyond professional engineering departments, while the resulting applications gained the ability to make decisions and act. Zenity's move into agent security is an extension of the founders' original thesis rather than a late pivot toward generative AI.

Zenity's current platform is designed to inventory agents, inspect their configurations and permissions, monitor runtime behavior, and prevent actions that violate policies. Zenity lists support for agent environments spanning Microsoft Copilot, ChatGPT Enterprise, Claude Enterprise, Salesforce Agentforce, ServiceNow, AWS Bedrock, Google Vertex AI and custom systems.

Kliger said in the funding announcement that he and Bargury founded Zenity around the belief that software development was no longer reserved for skilled engineers. The financing gives Zenity the resources to pursue the consequence of that shift: every worker or department that can deploy an agent can also create a new identity, data-access path and execution surface that security departments must govern.

Norwest backs the agent as a new security perimeter

Norwest partner Assaf Harel framed the investment around the inadequacy of security controls built for human activity. In Norwest's investment memo, Harel argued that security information, identity, data-loss prevention and endpoint products can record individual events while missing the intent connecting an agent's multi-step activity.

That distinction is central to Zenity's pitch. A compromised agent can have legitimate credentials and approved access while still performing a harmful sequence of actions. Zenity says its software correlates how an agent was built and configured with what it attempts at runtime, then allows, modifies or blocks the action before completion.

Norwest said its diligence included conversations with security executives and prospective customers. The investor also cited customer results published by Zenity, including one Fortune 20 technology customer that Zenity says remediated 90% of identified vulnerabilities while its agent deployment grew 280% over a year. Those performance figures come from Zenity and do not disclose the customers' identities or the starting number of agents.

Zenity says most of its customers are Fortune 500, Global 2000 and other large organizations, naming SoftBank Corp. as one customer. Zenity also says revenue tripled in each of the past two years and is on course to triple again in 2026. Zenity has not published revenue, annual recurring revenue or a customer count, leaving the scale behind those growth rates unclear.

The round finances a global buildout

Zenity plans to use the Series C for product development, research and international expansion, particularly in Europe and the Asia-Pacific region. Zenity says it employs more than 230 people, with research and development centered in Tel Aviv and commercial operations led from New York.

The headcount marks a rapid expansion from the 52 employees reported around the Series B in 2024. Zenity now has to convert that hiring and the new capital into repeatable enterprise deployments while large security vendors, specialist AI-security providers and cloud platforms pursue the same control point.

Zenity's founders have one structural advantage: they entered the market through low-code governance, giving them experience with software created outside conventional engineering workflows. The Series C tests whether that head start can translate into a durable security layer across competing agent platforms, or whether cloud and application providers will absorb the controls into their own products.

The new investors also reflect that distribution challenge. SoftBank Vision Fund 2 brings a connection to a named Zenity customer, while Hitachi Ventures and LG Technology Ventures provide ties to large corporate technology buyers in Asia. Norwest's Harel focuses on multi-stage investments in Israeli-related enterprise software companies, aligning the lead investor with Zenity's Tel Aviv research base and New York commercial operation.

For Kliger and Bargury, the funding turns a governance thesis formed inside Microsoft into a much larger bet: enterprises will deploy agents faster than their existing security architecture can adapt, and they will pay for an independent control layer capable of following those agents across vendors, clouds and employee devices.