Zenity Raises $125 Million for AI Agent Security
Zenity raised $125 million in a Series C round led by Norwest on August 3, 2026, to expand its AI agent security business. New investors include Qumra Capital, SoftBank Vision Fund 2, Hitachi Ventures, and LG Technology Ventures, alongside existing backers. According to Zenity, the funding supports global expansion, platform innovation, and enterprise demand for controls over autonomous AI-agent actions.
Zenity raised $125 million in a Series C funding round led by Norwest, the AI security and governance vendor announced on August 3, 2026. Qumra Capital, SoftBank Vision Fund 2, Hitachi Ventures, and LG Technology Ventures joined the round, alongside existing investors Vertex Ventures, Third Point Ventures, DTCP, and Intel Capital.
According to Zenity's announcement, the capital is intended for global expansion, platform innovation, and meeting enterprise demand for AI-agent security. Fortune reported that the Tel Aviv-based cybersecurity startup focuses on securing AI agents operating inside corporate systems.
Security controls for agent actions
The funding arrives as enterprises deploy agents that can invoke tools, access connected systems, make decisions, and execute actions. Fortune reported that Zenity's approach differs from controls focused chiefly on prompt screening or post hoc output auditing: its platform monitors agent actions in real time and can block or alter actions that deviate from an agent's original purpose.
Zenity describes its approach as interpreting an agent's intent and deterministically permitting, altering, or blocking an action before it occurs. FinTech Global reported that the platform covers agentic deployments across Microsoft Copilot, ChatGPT Enterprise, Gemini, Claude, Codex, Cursor, AWS Bedrock and AgentCore, Microsoft Foundry, and Google Vertex AI.
That distinction matters technically because agentic systems expand the attack surface beyond a model response. A compromised or poorly bounded agent may inherit permissions from SaaS applications, cloud services, code repositories, and other tools. In comparable enterprise deployments, practitioners generally need controls that can evaluate identity, authorization, tool invocation, data access, and execution context, rather than treating the prompt and response as the complete security boundary.
Enterprise adoption and reported growth
Zenity stated that most of its customers are Fortune 500, Global 2000, and other large global organizations, including SoftBank Corp., and that it serves regulated sectors including financial services, healthcare, pharmaceuticals, energy, and manufacturing. The company also reported that revenue tripled in each of the past two years and that it is on track to triple revenue again this year. Those growth figures are company-reported and were not independently verified in the available sources.
SoftBank Corp. Senior Vice President and CISO and CRO Tadashi Iida said in Zenity's release: "Zenity enables us to confidently deploy AI agents across the enterprise, giving us the visibility and governance required to support innovation at scale."
Norwest partner Assaf Harel characterized Zenity as an early mover in end-to-end AI-agent security and said the company has a growing footprint among Fortune 1000 customers. Those are investor assessments rather than independent market measurements.
FinTech Global also reported that Gartner named Zenity a frontrunner in AI agent governance in an April 2026 report, citing agentic-centric architecture and intent-aware detection. The underlying Gartner report was not included in the retrieved materials.
What practitioners should watch
The round is notable for placing agent governance, not only model security, at the center of an AI-security investment thesis. Across the sector, the emergence of agents with delegated permissions has increased attention on runtime authorization, tool-use policies, audit trails, and controls against prompt injection or indirect instruction attacks.
FinTech Global reported that Zenity Labs has published research on vulnerabilities including AgentFlayer, described as a class of zero-click attacks targeting enterprise agents, as well as findings involving Microsoft Copilot Studio, Perplexity's Comet browser, and connected ChatGPT documents. For teams evaluating agent-security platforms, independently testing coverage across their own identity stack, tool protocols, data connectors, and agent frameworks remains important because enforcement guarantees depend on where the control plane sits in the execution path.
Key Points
- 1Zenity's $125 million Series C funds an agent-security platform as enterprises grant AI systems authority to access tools and corporate systems.
- 2Reported platform controls target runtime agent actions, extending security evaluation beyond prompts and model outputs to authorization and execution context.
- 3Across comparable deployments, delegated AI-agent permissions increase the importance of auditable tool-use policies, identity controls, and runtime enforcement.
Scoring Rationale
A $125 million round is a notable investment in the emerging AI-agent security and governance market. It is relevant to ML and security practitioners because agentic deployments introduce runtime tool-use, identity, and authorization risks beyond prompt-level controls.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
