Zenity AI runtime security
Zenity concentrates on agents and says it stops harm in the middle of a chain. Boundaries stops multi-step exfiltration and privilege-escalation chains before they complete, and native hooks plus an MCP gateway can block or modify a dangerous tool call, including malicious package installs and network calls to listed domains. Its AI detection and response tracks each agent step in one activity table, and detections map to OWASP LLM and MITRE ATLAS. Third-party press confirms its funding and research. Gaps from 16 Nov 2026: inline latency, fail-open or fail-closed behavior, a price and documentation readable without sign-in are none found, since the docs sit behind a login.
Facts read on .
Strength and weakness for AI Runtime Security Tools
Strength: Documents blocking of tool calls, multi-step exfiltration chains, prompt injection and malicious package installs in real time across agent and endpoint-hook layers (zenity-46, zenity-50, zenity-52, zenity-51).
Weakness: Published latency and fail-open or fail-closed behavior are none found on 2026-11-16, and docs are login-gated (zenity-56, zenity-39).
Scores and evidence
Zenity is 12th in the overall ranking at 50.0 of 100. Each score below lists the fact ids behind it; the facts follow further down with their sources and read dates.
| Criterion | Score | Weight | Why | Facts |
|---|---|---|---|---|
| Action types stopped before they complete | 7 of 10 | 30 | The vendor says it blocks tool calls, prompt injection, data leakage, secret exfiltration, malicious package installs and network calls to listed domains before execution, all on vendor claims. Blocking of harmful output as a separate control: none found. | zenity-24, zenity-50, zenity-51, zenity-52, zenity-53, zenity-48, zenity-38, zenity-57 |
| Layers covered | 6 of 10 | 20 | The vendor claims enforcement across agent tool calls, AI applications and coding agents on developer machines. Endpoint install details: none found, and all of it is vendor-only. | zenity-45, zenity-47, zenity-20, zenity-50, zenity-29 |
| Latency and failure mode | 0 of 10 | 15 | Published inline latency and fail-open or fail-closed behavior: none found. | zenity-56 |
| Deployment | 4 of 10 | 15 | The vendor describes API integrations, native hooks and an MCP gateway URL. Operating systems, install method, time to deploy and readable docs: none found. | zenity-25, zenity-26, zenity-37, zenity-29, zenity-58, zenity-39 |
| Detection and response | 6 of 10 | 5 | The vendor claims a live view, audit trails, an agent activity table and rule preview, and its docs navigation lists Cribl, CrowdStrike and Palo Alto integrations. Integration detail: none found, because the docs are behind sign-in. | zenity-27, zenity-28, zenity-54, zenity-49, zenity-37 |
| Transparency | 4 of 10 | 10 | Third-party press confirms the vendor's published research. Documentation readable without sign-in and a published price: none found. | zenity-12, zenity-13, zenity-14, zenity-39, zenity-59 |
| Proof | 8 of 10 | 5 | Third-party press confirms the funding, and the vendor names a customer and cites Gartner, Latio and SACR coverage. The analyst coverage is vendor-reported. | zenity-05, zenity-02, zenity-04, zenity-08, zenity-06 |
Where it ranks
Product, deployment, price and owner
- Product assessed: Zenity Boundaries (runtime enforcement engine) and AI Detection & Response (AIDR), with native agent hooks and the Zenity MCP gateway for coding agents
- Deployment: Connects to platforms by API and integration (Microsoft, Salesforce, OpenAI, Google, AWS, Okta and others per docs navigation); coding agents via native hooks, OpenTelemetry and the Zenity MCP gateway; no agent or proxy install details published
- Time to deploy: none found on 2026-11-16
- Price: none found on 2026-11-16
- Owner: Zenity (standalone, venture-backed company; no parent company stated on its pages)
- Website: zenity.io
Facts and sources
zenity-01 (vendor): Zenity describes itself as an AI agent security platform that spans discovery, policy, identity, and runtime defense.
Quote: "Zenity secures long-horizon AI agents at the decision, the exact point where context and intent come together to create the highest risk."
Source: zenity.io, read .
zenity-02 (vendor): On 2026-08-03, Zenity announced a $125 million Series C led by Norwest; it is a standalone venture-backed company headquartered in New York.
Quote: "today announced a $125 million Series C led by Norwest."
Source: zenity.io, read .
zenity-03 (vendor): Zenity lists its funding history on its company page: a $5M seed in 2021, a $16.5M Series A in 2023 and a $38M Series B in 2024.
Quote: "Zenity raises $38M Series B round of fundraising, co-led by Third Point Ventures and DTCP"
Source: zenity.io, read .
zenity-04 (vendor): Zenity says most of its customers are Fortune 500 and Global 2000 organizations and names SoftBank Corp. as a customer.
Quote: "including SoftBank Corp."
Source: zenity.io, read .
zenity-05 (third-party): Israeli business outlet Calcalist CTech, on 2026-09-03, reports that Zenity recently raised $125 million in a round led by Norwest.
Quote: "a company that recently raised $125 million in a funding round led by Norwest"
Source: calcalistech.com, read .
zenity-06 (vendor): On 2026-09-15, Zenity announced recognition in two independent analyst reports, from Latio and Software Analyst Cyber Research (SACR).
Quote: "today announced recognition in two independent analyst reports examining the evolving AI agent security market."
Source: zenity.io, read .
zenity-07 (vendor): Zenity says SACR places it among vendors combining runtime behavior analysis, MCP and tool governance, agent identity, audit evidence and response.
Quote: "SACR identifies Zenity among vendors that combine multiple agent security capabilities, including runtime behavior analysis, MCP and tool governance"
Source: zenity.io, read .
zenity-08 (vendor): On 2026-09-23, Zenity announced it was named a Market Shaper in Gartner's first Emerging Market Quadrant for AI Application Security (startup vendors).
Quote: "Zenity has been named a Market Shaper in Gartner"
Source: zenity.io, read .
zenity-09 (vendor): Zenity lists a Gartner paper naming it the "Company to Beat" in AI agent governance on its platform pages.
Quote: "GartnerĀ® Names Zenity as the Company to Beat in AI Agent Governance"
Source: zenity.io, read .
zenity-10 (vendor): On 2026-01-06, Zenity announced it is available on AWS Marketplace with support for Amazon Bedrock AgentCore, and says its coverage spans cloud, SaaS, endpoints and agentic browsers.
Quote: "across cloud, SaaS, endpoints, and agentic browsers"
Source: zenity.io, read .
zenity-12 (third-party): On 2026-09-24, The Register reported that Zenity Labs found three Salesforce Agentforce flaws (SalesBleed) and reported them to Salesforce, which fixed them.
Quote: "Zenity Labs uncovered the three vulnerabilities, collectively called SalesBleed, and reported them to Salesforce"
Source: theregister.com, read .
zenity-13 (third-party): On 2026-09-25, SecurityWeek covered the SalesBleed Agentforce flaws that Zenity Labs reported.
Quote: "Three vulnerabilities in Salesforce Agentforce could have allowed attackers to hijack trusted agents for sensitive CRM data exfiltration and phishing, Zenity Labs reports."
Source: securityweek.com, read .
zenity-14 (vendor): On 2026-10-08, Zenity Labs published research on AgentCorruption, a chain of AWS AgentCore flaws.
Quote: "Zenity Labs Discloses AgentCorruption, a Chain of AWS AgentCore Flaws That Allowed One Prompt to Take Over All AgentCore Agents"
Source: zenity.io, read .
zenity-20 (vendor): Zenity names Claude Code and Cowork, ChatGPT Codex, GitHub Copilot and Cursor as the coding and personal agents where it enforces boundaries.
Quote: "enforces hard boundaries across Anthropic's Claude Code and Cowork, ChatGPT Codex, GitHub Copilot, and Cursor."
Source: zenity.io, read .
zenity-24 (vendor): Zenity says its MCP product can allow, modify or block a tool call based on live context.
Quote: "Allow, modify, or block a tool call based on live context, not a static allow-list."
Source: zenity.io, read .
zenity-25 (vendor): Zenity says MCP traffic is controlled by publishing a server and registering one gateway URL, so agents keep working unchanged.
Quote: "Publish a server, register one gateway URL, and agents keep working unchanged."
Source: zenity.io, read .
zenity-26 (vendor): Zenity says it collects coding-agent telemetry through native hooks and OpenTelemetry.
Quote: "through native hooks and OpenTelemetry: each step an agent takes, the code and data it touches"
Source: zenity.io, read .
zenity-27 (vendor): Zenity says its MCP Live View shows each request, tool invoked and identity behind it in real time.
Quote: "Live View shows each request, the tool invoked, and the identity behind it, in real time."
Source: zenity.io, read .
zenity-28 (vendor): Zenity says it provides unified visibility, audit trails and enforcement across coding and personal agents.
Quote: "Prove governance across every coding and personal agent, with unified visibility, audit trails, and enforcement."
Source: zenity.io, read .
zenity-29 (none-found): None found on 2026-11-16: operating systems supported, endpoint install method or time to deploy on developer machines, searched on the coding and personal agents page, MCP security page and platform page.
Source: zenity.io, read .
zenity-35 (vendor): Zenity says it uses identity to gate agent actions, such as blocking a deactivated Okta user or restricting privileged tools by role.
Quote: "Zenity uses identity to gate what an agent can do, like blocking a deactivated Okta user or restricting privileged tools by role."
Source: zenity.io, read .
zenity-36 (vendor): Zenity says it flags risky non-human identities such as static API keys that coding agents use and that never appear in the IdP.
Quote: "Surface risky NHIs, including static API keys and secrets that coding agents use and that never appear in your IdP."
Source: zenity.io, read .
zenity-37 (vendor): Zenity's public docs navigation lists integrations for OpenAI, Google, AWS, Salesforce, Anthropic, Cursor, GitHub, LiteLLM, Cribl Stream, Okta, CrowdStrike and Palo Alto Networks.
Quote: "Gateways and custom agents LiteLLM Custom agents Cribl Stream ACS schema Identity and endpoint security Okta CrowdStrike Palo Alto Networks"
Source: docs.zenity.io, read .
zenity-38 (vendor): Zenity's docs navigation lists pages for Boundaries, Inline Prevention, Kill switch, Malicious Package Protection and AI Detection & Response.
Quote: "Protect Incidents AI Detection & Response Malicious Package Protection Enforce Boundaries Boundaries demos Boundaries CLI Inline Prevention Kill switch"
Source: docs.zenity.io, read .
zenity-39 (none-found): None found on 2026-11-16: documentation body text readable without sign-in; docs.zenity.io/user-guide/* and /release-notes redirected to a login page, and only the navigation tree of the docs home page loaded.
Source: docs.zenity.io, read .
zenity-42 (vendor): Zenity says inventory appears the moment a platform is connected, with no separate discovery step.
Quote: "Full inventory the moment a platform connects, no separate discovery step required"
Source: zenity.io, read .
zenity-45 (vendor): Zenity says its Boundaries engine evaluates every agent action in real time and can let it through, block it, or shut the agent down.
Quote: "evaluates every agent action in real time against rules of an organization and decides to let it through, block it, or shut the agent down."
Source: zenity.io, read .
zenity-46 (vendor): Zenity says Boundaries stops multi-step exfiltration and privilege-escalation chains before they complete.
Quote: "Multi-step exfiltration and privilege-escalation chains get stopped before they complete, not flagged after the fact."
Source: zenity.io, read .
zenity-47 (vendor): Zenity says one rule is enforced across Copilot, ChatGPT Enterprise, Agentforce, Bedrock, coding agents and homegrown builds.
Quote: "Write a rule once and it's already enforced on Copilot, ChatGPT Enterprise, Agentforce, Bedrock, coding agents, and homegrown builds"
Source: zenity.io, read .
zenity-48 (vendor): Zenity gives an example of one rule blocking tool calls to specific third-party domains across every agent.
Quote: "A global asset manager blocks tool calls to specific third-party domains across every agent with one rule"
Source: zenity.io, read .
zenity-49 (vendor): Zenity says a rule can be previewed to show what it would have caught before it can block anything.
Quote: "See exactly what a rule would have caught before it can block anything"
Source: zenity.io, read .
zenity-50 (vendor): Zenity says that through native agent hooks and the Zenity MCP gateway it can block or modify a dangerous tool call before it executes.
Quote: "it can block or modify a dangerous tool call before it executes."
Source: zenity.io, read .
zenity-51 (vendor): Zenity names secret exfiltration, malicious package installs and unsafe tool chaining as risks it detects and prevents on coding agents.
Quote: "Detect and prevent risks like secret exfiltration, malicious package installs, unsafe tool chaining, and an agent working around a security control"
Source: zenity.io, read .
zenity-52 (vendor): Zenity says AIDR detects direct and indirect prompt injection and blocks execution before the agent acts on it.
Quote: "AIDR detects direct and indirect prompt injection and blocks execution before the agent acts on them."
Source: zenity.io, read .
zenity-53 (vendor): Zenity says AIDR monitors and blocks sensitive data leaving through agent conversations and tool calls.
Quote: "Real-time monitoring and blocking of sensitive data leakage leaving through agent conversations, tool calls, or encoded payloads"
Source: zenity.io, read .
zenity-54 (vendor): Zenity says AIDR tracks each agent step, including triggers, retrievals, tool calls and agent handoffs, in one activity table.
Quote: "Step-by-step tracking across triggers, retrievals, tool invocations, and agent handoffs."
Source: zenity.io, read .
zenity-55 (vendor): Zenity says AIDR detections map to OWASP LLM and MITRE ATLAS and can block unsafe actions in real time.
Quote: "detecting threats mapped to the OWASP LLM and MITRE ATLAS frameworks and blocking unsafe actions in real time."
Source: zenity.io, read .
zenity-56 (none-found): None found on 2026-11-16: published latency of an inline decision, or published fail-open or fail-closed behavior and timeouts, searched on the Boundaries, AIDR, MCP and coding agent pages.
Source: zenity.io, read .
zenity-57 (none-found): None found on 2026-11-16: documented blocking of harmful or toxic model output as a separate control, searched on the AIDR and Boundaries pages (data leakage and prompt injection are documented).
Source: zenity.io, read .
zenity-58 (none-found): None found on 2026-11-16: a published time to deploy for the platform, searched on the platform page, product pages and press releases.
Source: zenity.io, read .
zenity-59 (none-found): None found on 2026-11-16: a published price or pricing page; https://zenity.io/pricing returned HTTP 404 and the sitemap lists no pricing page.
Source: zenity.io, read .