Zenity raised $125 million in a Series C on August 3, 2026, bringing its total funding to $185 million and positioning it as the best-capitalized pure-play AI agent security company in enterprise technology. The round was led by Norwest Venture Partners, with SoftBank Vision Fund 2, Hitachi Ventures, and LG Technology Ventures joining as new investors. Gartner named Zenity “The Company to Beat in AI Agent Governance” in April 2026. The funding arrives as enterprise AI agent deployments accelerate past the experimentation phase and the question of how to secure autonomous software becomes urgent.
The Problem AI Agent Security Is Solving
Traditional enterprise security was built for two types of software: applications that do exactly what a developer instructed, and humans who sometimes make mistakes. AI agents fit neither category. They operate autonomously, initiate multi-step actions without a human in the loop at each decision point, access production systems and sensitive data, and can be manipulated mid-workflow by adversarial inputs they encounter while executing a task.
The attack surface is new and growing fast. According to Gartner, the average Fortune 500 will operate more than 150,000 AI agents by 2028. Most enterprises currently run AI agents on governance models inherited from traditional application security: standing API keys, service account tokens, and post-incident log review. Those controls were not designed for software that reasons and acts autonomously.
Zenity CEO Ben Kliger summarized the shift in the company’s official announcement: “AI experimentation is long over. Any organization on the planet is promoting AI agents at velocity and adoption rates never seen before in any tech wave. They are rebuilding their infrastructure with agentic infrastructure and employees that access enterprise data, tools, and own business processes and decisions.”
The risk is not theoretical. Two major AI research labs disclosed in late July 2026 that their models had breached systems outside their intended evaluation scope. If models built and monitored by frontier labs can escape boundaries during controlled testing, the question of what happens to the hundreds of agents running unsupervised inside enterprise environments becomes very concrete.
What Zenity Builds: Intent-Aware, Pre-Execution Security
Zenity’s core architectural decision is to evaluate agent intent before action, rather than inspecting prompts or reviewing logs after the fact. The platform analyzes what an agent is trying to do in the context of its assigned purpose, then deterministically allows, modifies, or blocks the action before it executes.
That distinction matters at scale. An AI agent processing invoices should never need to query employee compensation data. If an agent making a routine API call suddenly attempts to read a sensitive system configuration file, that is not a prompt injection pattern to scan for: it is a behavioral anomaly that only makes sense when you understand what the agent is supposed to be doing and compare it to what it is actually attempting. Zenity’s intent model does that comparison continuously, in real time, for every agent action.
The platform’s key capabilities, as described in the company’s announcements and reported by Fortune on August 3, 2026:
| Capability | What It Does | Why It Matters |
|---|---|---|
| Intent detection | Evaluates agent purpose before action executes | Catches novel manipulation that rule-based scanning misses |
| Unified coverage | Works across all major agent frameworks | Eliminates gaps from shadow deployments or mixed stacks |
| Real-time enforcement | Approves, modifies, or blocks before execution | Prevents incidents rather than responding to them |
| Continuous monitoring | Tracks agent behavior across entire workflows | Identifies drift from intended behavior over time |
| Research-backed rules | Zenity Labs discloses vulnerabilities before rules ship | Policies reflect real-world attack patterns, not lab scenarios |
Zenity integrates across every major enterprise AI environment: Microsoft Copilot, ChatGPT Enterprise, Gemini, Claude Code, Codex, Cursor, custom-built agents in AWS Bedrock and AgentCore, Microsoft Foundry, and Google Vertex AI. That breadth is the product requirement. Enterprises do not run a single agent on a single platform. They run dozens of agents across every major cloud vendor, every major SaaS tool, and internal systems. A security layer that covers only one vendor creates a governance gap that adversaries immediately exploit.
The Investor Lineup and What It Signals
The composition of Zenity’s Series C is as significant as the size. SoftBank Vision Fund 2 participation is notable because SoftBank Corp, the operating company, is already a paying Zenity customer. Tadashi Iida, Senior Vice President and CISO at SoftBank Corp, confirmed in the announcement: “Zenity enables us to confidently deploy AI agents across the enterprise, giving us the visibility and governance required to support innovation at scale.” The Vision Fund investing in a vendor its own operating company already relies on for AI agent governance is a strong signal about production readiness.
Hitachi Ventures and LG Technology Ventures represent a different signal: the industrial and manufacturing sector’s emerging concern about autonomous AI in operational environments. Both companies operate in sectors where AI agents increasingly touch physical processes, supply chains, and compliance-sensitive workflows. Their participation reflects a thesis that AI agent security is not a software problem confined to technology companies. It is an infrastructure problem for any large organization deploying agents at scale.
Norwest leading the round, with returning investors Vertex Ventures, Third Point Ventures, DTCP, and Intel Capital, brings the total capital raised to $185 million. According to Calcalis Tech, Zenity has tripled annual revenue in each of the past two years and is on pace to triple again in 2026.
Who Founded Zenity
Zenity was founded in 2021 by Ben Kliger (CEO) and Michael Bargury (CTO), both veterans of Israeli military intelligence unit 8200 who previously worked together at Microsoft building cloud and operational technology security products. That background is directly relevant. Their Microsoft work involved securing infrastructure where the difference between legitimate access and unauthorized behavior is often subtle and context-dependent. The same problem appears in AI agent governance, where an agent holding valid credentials can still be operating outside its intended scope.
Zenity Labs, the company’s research arm, has disclosed multiple high-profile vulnerabilities affecting enterprise AI systems, including attacks targeting Microsoft Copilot Studio and Perplexity’s Comet browser agent. That research feeds directly into the platform’s enforcement rules, which means Zenity’s policies reflect real-world adversarial techniques rather than theoretical attack models.
The company now employs more than 230 people worldwide, with R&D in Tel Aviv and go-to-market operations led from New York.
What This Round Means for Enterprise AI Strategy
Zenity’s $125M round is the third significant fundraise in the AI agent governance category in the past 45 days. Hush Security raised $30M in late July 2026 to build just-in-time credential infrastructure for AI agents. Snowflake’s Cortex AI Gateway added agent control plane capabilities to its data platform the same week. The pattern is consistent: enterprises are actively procuring governance infrastructure, and investors are responding to that demand signal.
The convergence of capital in this category reflects a structural reality. Enterprise AI agent deployments have crossed a threshold. Organizations are no longer running one or two agents in a controlled environment. They are deploying agents across Salesforce, Workday, ServiceNow, Slack, GitHub, and internal systems simultaneously. The attack surface is as broad as the enterprise technology stack itself. Security tools that were not built for this environment are not going to cover it.
For enterprise AI leaders, the Zenity round is a useful calibration point. If a company backed by SoftBank, Hitachi, and LG with Fortune 500 customer endorsements and Gartner recognition is raising $125M to solve AI agent security, the implication is that most organizations are not yet adequately covered. The question is not whether to govern AI agents but whether to build that infrastructure internally or adopt a platform purpose-built for the problem.
Enterprises already running agents across multiple AI workspaces and agentic platforms can book a strategy session with Enera at /book-a-call to assess current governance posture and identify the gaps most likely to create operational or compliance exposure.
The Era of 1 Billion AI Agents
The headline framing in Zenity’s announcement is “the era of 1 billion AI agents.” The number is directional, not a specific forecast. The point is that AI agents are no longer a pilot project. They are production infrastructure, and the governance models enterprises used during the pilot phase were not designed to scale.
Zenity is betting that the tools and architecture required to secure that infrastructure at scale represent a category-defining opportunity. The investor roster suggests that bet has significant institutional backing. The customer list, anchored by Fortune 50 companies in regulated industries, suggests the market is willing to pay for it today.
The Series C funding will accelerate platform innovation, expand Zenity Labs, and deepen the company’s presence across Europe and Asia Pacific. According to the company’s press release, the investment is also intended to help define the security standards enterprises rely on as AI agents become foundational to modern business. That last ambition, standard-setting, is the signal that Zenity sees this not just as a product race but as an infrastructure category it intends to shape.