Zenity secures and governs AI agents in the enterprise. Security teams use it to see which agents are running, check what each one is allowed to touch before it goes live, and decide exactly what agents can and cannot do once they are acting on company data and systems.
The problem it addresses is that agents now arrive from everywhere at once: built into SaaS suites, installed by employees on their own machines, and built in-house on cloud AI services. Few security teams can say how many exist, let alone what each one may do.
At A Glance
| Vendor | Zenity |
|---|---|
| Category | AI agent security and governance |
| Identities covered | AI agents across SaaS, endpoint, cloud and homegrown deployments |
| Deployment | SaaS platform connected to enterprise AI platforms |
| Agentic AI | Agent inventory, posture management, runtime boundaries, Agentic Identity, MCP Security |
| Certifications | SOC 2 Type II, ISO 27001, ISO 27701 |
| Funding | Series C (2026) |
| Market classification | Agentic AI Identity; also NHI Platforms & Machine Identity |
What Zenity Does
Once an AI agent can read mailboxes, update records or call other tools, a bad instruction or an over-generous permission becomes a real business risk. Zenity is built to govern how agents behave rather than only how they are configured. Its controls follow an agent decision from before it happens, through the moment it is made, to what follows: an agent’s access is checked ahead of launch, its intended action is interpreted at runtime and then allowed, changed or blocked, and anything that slips through is detected and answered. One policy model applies whether the agent lives in a SaaS suite, on an endpoint or in a cloud account.
Products
AI Observability
AI Observability answers the first question most security teams have: which agents exist and what data they reach. It discovers agents across SaaS, custom and endpoint deployments and keeps the inventory current as new ones appear.
AISPM
AI Security Posture Management checks each agent’s configuration and permissions against policy before it goes live, so over-broad access is caught at build time rather than after an incident.
Exposure Management
Exposure Management sorts theoretical risk from real risk. It validates which attack paths against an agent can actually be exploited, scores each one and produces a fix that can be applied in Boundaries.
Boundaries, AIDR And Further Modules
Boundaries is where runtime limits on agent behaviour are set, including the fixes that come out of Exposure Management. AIDR adds detection and response for AI agents, with lightweight monitoring for personal and coding agents. Two further modules deal with agent-specific risks: Agentic Identity covers the identities agents act under, and MCP Security covers the MCP connections agents use to reach tools.
How It Works
Zenity runs as a SaaS platform that connects to the enterprise AI services where agents live, including Microsoft, Salesforce, ServiceNow, AWS, Google and OpenAI products. Inventory and posture data are read from those platforms through their own connections, runtime decisions are made as each agent action is evaluated, and fixes found during exposure testing flow into the runtime limits. Personal and coding agents such as Cursor, which run on employee machines rather than in a managed platform, are covered by the lighter monitoring layer instead. Customers can buy the platform through Azure Marketplace or AWS Marketplace.
Identities And Systems Covered
| AI agents | SaaS-embedded, personal, coding, cloud-hosted and custom-built agents |
|---|---|
| Microsoft | Microsoft 365 Copilot, Copilot Studio, Power Platform and Microsoft Foundry |
| Cloud AI platforms | Amazon Bedrock, Amazon Bedrock AgentCore and Google Vertex AI |
| SaaS AI | ChatGPT Enterprise, Claude Enterprise, Salesforce Agentforce and ServiceNow |
| Coding agents | Cursor |
Company Background
| First launch | 2021 |
|---|---|
| Headquarters | Tel Aviv, Israel, with go-to-market and operations run from New York |
| Leadership | Ben Kliger, Co-Founder and CEO; Michael Bargury, Co-Founder and CTO |
| Funding | $5 million seed (2021), $16.5 million Series A led by Intel Capital (2023), a strategic investment from M12 (2024), $38 million Series B (2024) and $125 million Series C led by Norwest (August 2026) |
| Recognition | Included in Gartner Cool Vendors in Agentic AI TRiSM and in Gartner’s Emerging Market Quadrant for AI Application Security |
Note: This overview is compiled from Zenity’s publicly available information as of 3 November 2026. It is provided for information only. NHI Mgmt Group does not endorse any vendor, and nothing on this page represents the opinions of NHI Mgmt Group.