All analysis was generated autonomously, without human review. Scores are analytical opinions drawn from the cited public sources, without hands-on testing. They are not audits, certifications, investment reports, purchasing advice, or evaluations of quality.
Zenity sells large enterprises a security platform for their AI agents. It inventories agents, checks their permissions before launch, and allows or blocks each agent action under one rule set across Copilot, ChatGPT Enterprise, Agentforce, and Bedrock. Founded in 2021, Zenity first secured low-code apps, for customers including Varonis and Telit Cinterion. It also names SoftBank Corp and WM as customers and says most customers are Fortune 500, Global 2000, or other leading global organizations. Norwest led Zenity's $125 million Series C, which SecurityWeek says brings its total to $180 million. That rule set suits enterprises running agents on several platforms. If platform owners such as Microsoft build comparable controls into their agents, Zenity would rely more on those enterprises.
| Description | Zenity sells a security platform that governs how enterprise AI agents behave, with agent inventory, posture checks, runtime detection, and enforcement of rules on agent actions. | [f1] |
|---|---|---|
| Founded | 2021 | [f2] |
| HQ | Tel Aviv, Israel | [f2] |
| Funding | $180M total | [f3] |
| Latest funding | Series C, $125M led by Norwest (August 2026) | [f3] |
| Deployment | SaaS | [f4] |
| Compliance | GDPR, ISO 27001, ISO 27701, SOC 2 Type 2 | [f4] |
| Product | What it does |
|---|---|
| Zenity | Zenity: Secures enterprise AI agents with discovery, posture management, and runtime detection and response across agent platforms. |
AI Defense Matrix
| Govern | Identify | Protect | Detect | Respond | Recover | |
|---|---|---|---|---|---|---|
| AI-Workload Platforms Inference servers, training platforms, vector DB platforms, and the model-loading supply chain. | ||||||
| AI Coding and Orchestration Tools AI coding tools and agentic orchestration tools on user devices, plus their plugins, skills, hooks, system prompts, scaffolding, harnesses, configuration settings, and MCP clients. | ||||||
| AI-Generated Code Code produced by AI tools, AI-assisted reviews, AI-generated infrastructure-as-code and tests, and vibe-coded apps that bypass CI/CD. | ||||||
| AI Traffic Paths to external and self-hosted AI services, MCP tool channels, agent-to-agent calls, model-registry downloads, and egress to unapproved AI services. AI gateways, LLM routers, and MCP gateways steer traffic along those paths. The steering decisions and their inputs belong here too: routing policies, MCP server registries, and agent naming and discovery services. | ||||||
| AI Model Model weights, fine-tuning checkpoints, model cards, registries, an AI bill of materials (AIBOM), and the third-party LLMs your enterprise consumes. | ||||||
| Training Data Datasets used for training, fine-tuning, and continued learning. | ||||||
| Runtime AI Data User prompts, inference inputs, RAG content, vector DB content, persistent agent memory, and interaction history. | ||||||
| AI Agent Identities AI agents as non-human principals, plus credentials, keys, permission scopes, service accounts, and delegation chains across agents and tools. |
Zenity secures enterprise AI agents with discovery, posture management, and runtime detection and response across agent platforms. It is mapped to the AI Defense Matrix. [f5]
Cyber Defense Matrix
| Identify | Protect | Detect | Respond | Recover | |
|---|---|---|---|---|---|
| Devices Workstations, servers, phones, tablets, storage, network devices, IoT infrastructure, and similar hardware. | |||||
| Applications Software, interactions, and application flows on the devices. | |||||
| Networks Connections and traffic flowing among devices and apps, plus communication paths. | |||||
| Data Content at rest, in transit, or in use across devices, apps, and networks. | |||||
| Users The people using the devices, apps, networks, and data. |
Zenity inventories low-code apps, automations, and flows built across Power Platform, surfaces risks such as hard-coded secrets, and enforces authentication and sharing policy. The Power Platform line is mapped to the Cyber Defense Matrix. [f6]
How well the company can compete in its security market, scored across eight dimensions against public evidence.
| Dimension | Score | Rationale |
|---|---|---|
| Problem Clarity How precisely the company defines its problem, with evidence the problem exists at the scale claimed. | 3/5 | Zenity names a specific buyer, the security team accountable for agents employees and developers run, and SecurityWeek, CSO Online, and The Register document the exposure through its researchers' findings. The figures on how much risk a buyer carries come from Zenity and its customers. The cited sources do not independently quantify that risk. [s4, s35, s36, s37, s16] |
| Capability Depth How specific the technical capabilities are, with evidence beyond marketing claims such as docs and third-party validation. | 3/5 | Zenity documents concrete mechanisms on its own pages: AIDR breaks every agent interaction into steps from user messages to tool calls and retrievals, the Boundaries engine flags rules that duplicate or contradict existing ones, and the MCP gateway lets a customer publish a server behind one gateway URL. The outside coverage reports funding and Zenity’s attack research on other platforms, and the analyst readings reach the record through Zenity’s releases, so no independent technical evaluation of Zenity’s controls appears. [s5, s6, s7, s14, s36] |
| Market Timing Whether the market is ready for this product, with evidence that buyers are actively seeking solutions. | 3/5 | The enabler is the spread of tool-using AI agents inside enterprise platforms, which an April 2026 Gartner report quoted by Zenity describes as invigorating activity in the AI agent governance market. Zenity's 2026 announcements quote analyst assessments from Gartner, Latio, and SACR and include WM's account of operating agents at enterprise scale. The cited sources do not independently document multiple kinds of buyer demand. [s14, s15, s16, s17] |
| Team Credibility Demonstrated domain expertise with public signals such as prior exits, publications, and industry recognition. | 4/5 | Co-founder and CTO Michael Bargury leads research that CSO Online, The Register, and SecurityWeek have reported since 2025, from the AgentFlayer exploit chains at Black Hat USA to the SalesBleed flaws in Salesforce Agentforce. That sustained, multiply reported publication record meets the rung for recognized standing. [s8, s36, s37, s35, s33] |
| GTM Proof Evidence of actual traction (customers, revenue signals, partnerships) beyond stated intentions. | 4/5 | Security leaders at SoftBank Corp and WM speak for Zenity in its announcements, Varonis and Telit Cinterion appear in its case studies, and ServiceNow, AWS, and Carahsoft integrate, list, or resell the product. Revenue tripling in each of the past two years is Zenity's own figure, which Globes repeats, so scale is not independently corroborated. [s13, s16, s28, s29, s20, s23, s24, s19, s33] |
| Funding Efficiency Whether funding matches go-to-market ambition, with signs of capital-efficient growth. | 3/5 | Zenity has raised $180 million by SecurityWeek's count, including a $125 million Series C led by Norwest in August 2026, against a shipped platform, more than 230 employees, and named enterprise customers. The raise fits an enterprise sales motion, but its one growth figure, revenue tripling for two years, is the company's own, so output per dollar stays unconfirmed. [s32, s33, s34, s13, s2] |
| Category Clarity Whether the company creates or fits a recognizable category that buyers can quickly place in their stack. | 3/5 | Gartner placed Zenity under two framings in 2026, AI agent governance and AI application security, and named it in two Hype Cycle categories. Zenity reports these placements under different labels, so the category is recognizable but still contested. [s14, s15, s16, s17] |
| Incumbent Defensibility How vulnerable the core value proposition is to absorption as a feature by a platform vendor. | 3/5 | One Zenity rule applies across Copilot, ChatGPT Enterprise, Agentforce, Bedrock, coding agents, and homegrown builds, and ServiceNow's native SecOps integration adds workflow friction to removal. Salesforce builds its own controls, including the Trusted URLs filter SecurityWeek describes. A platform owner could add comparable controls for its own agents. [s6, s20, s35, s25] |
Enterprises now run AI agents that read data, call tools, and act inside business systems, and Zenity sells to the security teams accountable for them. Its own pages frame the risk as an agent's decision, where an outside prompt, an over-shared permission, or the agent's own improvisation leads it to act with access it should not use.
Independent reporting shows the exposure is real across major platforms. SecurityWeek reported that three SalesBleed flaws in Salesforce Agentforce could let attackers hijack trusted agents to exfiltrate CRM data. CSO Online covered Zenity's zero-click and one-click exploit chains affecting ChatGPT, Copilot Studio, Cursor, and Gemini. Salesforce fixed the flaws and reported no evidence that the issue had been exploited against any customer, so the findings document what attackers could do.
The figures on how large the problem is inside one enterprise come from Zenity's customers. WM's security head says Zenity surfaced policy violations across 575,000 resources and more than 200 environments, a customer statement published in Zenity's own announcement. [s1, s35, s36, s37, s16]
The Zenity platform covers an agent before, during, and after it acts. AI Observability builds a live inventory of agents and the data they touch, and AI Security Posture Management checks configuration, permissions, and integrations before an agent goes live. AI Detection and Response uses language models to read multi-step behavior and block unsafe actions in real time.
Boundaries is Zenity's runtime enforcement engine. It evaluates each agent action against an organization's rules and lets it through, blocks it, or shuts the agent down, and one rule applies across Copilot, ChatGPT Enterprise, Agentforce, Bedrock, coding agents, and homegrown builds. MCP Security uses the same engine to allow, modify, or block tool calls through a gateway URL.
Zenity still sells its original low-code line. The Power Platform offering inventories apps, automations, and flows and flags oversharing and hard-coded secrets. [s2, s3, s4, s5, s6, s7, s10, s28, s29]
Gartner's 2026 research gives Zenity standing in two framings. Zenity quotes an April report calling it the front-runner in AI agent governance. In September it announced a Market Shaper placement in Gartner's first Emerging Market Quadrant for AI Application Security, as one of two vendors in that quadrant among nearly 20 startups assessed.
Latio's 2026 report, as Zenity relays it, found the number of integrations Zenity supports high among the vendors it evaluated.
Platform owners are the other source of competition. Salesforce ships a Trusted URLs filter in Agentforce, and Zenity's SalesBleed research found that the filter missed some top-level domains and could be confused by certain characters. Salesforce then fixed the flaws, which shows both the opening for an outside control and the platform owner's capacity to harden its own agents. [s14, s15, s1, s17, s35]
ServiceNow, AWS, and Carahsoft each carry Zenity to enterprise or government buyers. ServiceNow SecOps natively includes Zenity security signals and controls, and Zenity says AWS selected it for the Security Hub Extended plan. AWS also lists Zenity on its Security Hub page, Zenity is available through AWS, and Carahsoft distributes it to agencies through contracts including NASA SEWP V.
Named customers now speak for the product. SoftBank Corp's CISO and WM's head of business information security are quoted in Zenity's releases, and Varonis and Telit Cinterion appear in its case studies. Zenity also publishes anonymized testimonials from Fortune 50 pharmaceutical and financial services companies on its homepage.
Zenity raised a $125 million Series C led by Norwest in August 2026, which SecurityWeek says brings its total to $180 million. Zenity says its revenue tripled in each of the past two years and that it has more than 230 employees, figures Globes and SecurityWeek repeat from the announcement. [s20, s22, s23, s24, s19, s13, s16, s28, s29, s1, s32, s33]
Ben Kliger, the CEO, and Michael Bargury, the CTO, co-founded Zenity in Tel Aviv in 2021 and still lead it. Zenity launched its low-code security platform that April and made it generally available in January 2022. Globes chose Zenity as one of its ten most promising startups of 2025.
Michael's research record is the team's most visible signal. CSO Online reported the AgentFlayer exploit chains Zenity presented at Black Hat USA, The Register covered an example he planned to demonstrate at RSA Conference, and SecurityWeek reported the team's SalesBleed findings in September 2026. [s8, s34, s36, s37, s35, s33]
Zenity's trust center lists SOC 2 Type II, ISO 27001, and ISO 27701 compliance and GDPR adherence, with full reports available on request. OpenCorporates records an active Israeli company under number 516350972 with the English legal name ZENITY.
Zenity announced FedRAMP In Process status in March 2026 and is pursuing authorization inside Knox Systems' existing authority to operate. The captured sources do not show that review completing, and Zenity said in March that it expected completion in upcoming review cycles. [s9, s38, s18]
| Company | Relationship | Note | Compare |
|---|---|---|---|
| Noma Security | competes with | Competes for the same enterprise budget for securing AI agents and applications. | |
| Prompt Security | competes with | Competes for enterprise AI-security budgets that span employee AI use and agents. | |
| Lasso Security | competes with | Competes for runtime protection of enterprise AI applications and agents. | |
| Lakera | competes with | Competes where buyers want runtime defenses for AI applications and agents. | |
| WitnessAI | adjacent | Adjacent in enterprise AI governance, where buyers weigh employee AI use alongside the agents Zenity secures. | N/AWe captured the evidence for these companies under different evidence-model versions (v1 vs v2), so the totals were scored under different conditions and are not directly comparable. |
| Knostic | adjacent | Adjacent where buyers govern what enterprise AI assistants can reach, close to Zenity's Copilot oversharing use case. | N/AWe captured the evidence for these companies under different evidence-model versions (v1 vs v2), so the totals were scored under different conditions and are not directly comparable. |
| Microsoft | adjacent | Owns Copilot Studio and Power Platform, surfaces Zenity protects, and invested in Zenity through its M12 fund. | N/AMicrosoft is scored by product line, not as a whole company, so there is no company-wide column to compare. Open its profile to compare a specific product. |
Add analyzed competitors to compare them side by side with Zenity.
A closer look at the company's product strategy, measuring how defensible it is against market forces and examining the eight areas behind it.
press the advantage
Zenity's Boundaries engine and inline runtime controls check agent actions before they run on Copilot, ChatGPT Enterprise, Agentforce, and Bedrock, so customers that route actions through them depend on Zenity in operation. Customers configure and run the software themselves. Zenity Labs publishes its agent research and offers a threat intelligence service for free, and its CTO says Zenity’s honeypots have captured attackers probing enterprise AI agents. Zenity aims at regulated enterprises and government agencies. Its US federal security authorization is under way, and the distributor Carahsoft offers it on public-sector contracts. Watch whether platform owners build comparable action-level controls into their own agents.
| Dimension | Score | Rationale |
|---|---|---|
| Value Delivery Does the product sell software as the product, or judgment, trust, or accountability with software as the delivery mechanism. | 1/3 | Customers buy Zenity's software and configure its inventory, posture checks, runtime rules, and playbooks themselves, and the reviewed product pages describe no service through which Zenity accepts accountability for the security outcome. |
| Switching Cost How expensive leaving is for a customer: data portability, integrations, learned workflows, network effects, regulatory data residency. | 2/3 | Leaving Zenity means rebuilding an agent inventory, posture policies, Boundaries rules, and the ServiceNow SecOps integration on another product, real friction in integrations and learned workflows. The cited record does not show those rules to be non-translatable or size what a migration would cost. |
| Compliance Moat Whether certifications, liability acceptance, or audit trails block an easy replacement. | 1/3 | Zenity lists SOC 2 Type II, ISO 27001, and ISO 27701 compliance and GDPR adherence, entry costs a funded rival can also meet. Its FedRAMP status is In Process through Knox Systems' authorization, so no authorization held by the product yet blocks a replacement. |
| Problem Complexity Whether the product requires ML, optimization, real-time systems, or years of specialized expertise. | 3/3 | Zenity uses language-model detections to read an agent's intent across multi-step chains of tool calls and retrievals and blocks unsafe actions in real time, and it applies that enforcement across Copilot, ChatGPT Enterprise, Agentforce, Bedrock, and coding agents. |
| Buyer Profile Whether buyers are SMB operators, mid-market IT teams, or regulated enterprises and governments with procurement gates. | 3/3 | Zenity is built and sold for regulated enterprises and government, with financial services and government offerings, FedRAMP In Process status, and Carahsoft distribution on contracts such as NASA SEWP V. Named buyers include SoftBank Corp and WM, and Zenity says the majority of its customers are Fortune 500, Global 2000, and other leading global organizations across financial services, healthcare, and other regulated industries. |
| Layer Whether the product is an end-user application, a platform with application features, or infrastructure other applications depend on. | 3/3 | Zenity's Boundaries engine and inline runtime controls sit in the path of agent actions, evaluating each action and letting it through, blocking it, or stopping the agent, and the Microsoft Foundry controls became generally available in March 2026. Zenity describes a financial services firm and an asset manager blocking agent messages and tool calls this way, so actions routed through those components wait on them. |
| Proprietary Data, Content, or IP Whether the product accumulates datasets, content licenses, or IP that a rival cannot recreate from scratch. | 2/3 | Zenity’s CTO told The Register that its global network of honeypots has captured attackers probing what they believe are legitimate enterprise AI agents, as prompt-level requests, and Zenity says its AI Total service records everything a submitted skill does at runtime. That is an accumulated, vendor-held record of attack activity, resting on Zenity’s own account, which a funded rival could build with comparable collection over time. |
Zenity sells to security teams at large and regulated enterprises that run AI agents across several platforms at once. The company says the majority of its customers are Fortune 500, Global 2000, and other leading global organizations in financial services, healthcare, pharmaceuticals, and other regulated industries. SoftBank Corp and WM speak for it by name.
Zenity also targets government and financial services buyers directly. On its financial services page Zenity promises the audit evidence regulated institutions need, on its government page it describes inline protection for agency agents, and Carahsoft distributes it to federal, state, and local agencies.
The segment grew out of low-code development. Zenity began in 2021 by securing apps and automations that business users build on platforms such as Power Platform, and it still serves that use, which Varonis and Telit Cinterion describe in its case studies.
Zenity's platform works in three layers that Zenity calls Surface, Enforce, and Protect. AI Observability and AI Security Posture Management inventory agents and check their configuration and permissions before they go live, and playbooks act on new violations automatically.
The runtime layers carry the AI-specific work. AI Detection and Response uses language-model detections to read intent across multi-step chains of tool calls and retrievals, mapped to OWASP and MITRE ATLAS. Boundaries evaluates each agent action against an organization's rules before letting it through, blocking it, or shutting the agent down.
Gartner's April 2026 report, quoted by Zenity, credits its intent-aware runtime defense and full-life-cycle observability. Zenity Labs also publishes agent attack research, including the AgentFlayer and SalesBleed findings that CSO Online and SecurityWeek reported.
Zenity sells through a demo-led enterprise motion supported by research and partners. On its homepage Zenity directs visitors to a risk assessment or a demo, and its Labs team publishes agent attacks that outlets including The Register, CSO Online, and SecurityWeek report.
Partners widen the channel. ServiceNow SecOps natively includes Zenity signals and controls, AWS lists Zenity on its Security Hub page, Zenity is available through AWS, and Carahsoft offers it to agencies through public-sector contracts.
The August 2026 Series C funds expansion. Zenity says the $125 million round led by Norwest will accelerate global expansion and platform work.
Zenity publishes no price on the pages reviewed, and on its homepage it directs visitors to a risk assessment or a demo request. A probe of zenity.io/pricing on 2026-10-20 returned a page-not-found response.
Buyers can also purchase Zenity through AWS and through Carahsoft's public-sector contracts.
Zenity connects to the platforms where agents live. Its documentation index lists integrations for Microsoft tenants, Salesforce, Google Vertex AI, ChatGPT Enterprise, Amazon Bedrock, Anthropic inference hooks, Cursor, and GitHub Copilot.
Some controls sit in the path of agent actions. Zenity made inline runtime controls for Microsoft Foundry agents generally available in March 2026, and its MCP Security gateway lets agents keep working while Zenity allows, modifies, or blocks each tool call.
Findings flow into tools security teams already run. ServiceNow SecOps ingests Zenity signals natively, and Zenity sends agent posture and runtime findings to AWS Security Hub in the OCSF format.
Zenity's trust center lists SOC 2 Type II, ISO 27001, and ISO 27701 compliance and GDPR adherence, with full reports available on request. OpenCorporates records an active Israeli company under number 516350972.
Federal authorization is in progress. Zenity announced FedRAMP In Process status in March 2026 and is pursuing authorization inside Knox Systems' existing authority to operate.
Zenity positions itself as one control point across many agent platforms. A single Boundaries rule applies on Copilot, ChatGPT Enterprise, Agentforce, Bedrock, coding agents, and homegrown builds, which suits enterprises that run agents on several platforms.
Zenity's inline controls for Microsoft Foundry came through an expanded partnership with Microsoft, whose M12 fund invested in Zenity in 2024.
Zenity Labs also shares tools with the wider community. Zenity offers a free threat intelligence service that runs AI agent skills in a contained environment and reports their behavior.
Ben Kliger, the CEO, and Michael Bargury, the CTO, co-founded Zenity and lead it, with research and development in Tel Aviv and go-to-market and operations run from New York. The company has more than 230 employees.
Michael's research gives the team its public profile. The Register quoted him explaining why most enterprise agents are open to zero-click attacks, and CSO Online and SecurityWeek reported the team's AgentFlayer and SalesBleed findings.
| Id | Source | Tier | Accessed |
|---|---|---|---|
| f1 | Zenity: platform overview page | official | 2026-10-20 |
| f2 | SecurityWeek: Zenity Raises $38 Million to Secure Agentic AI | press | 2026-10-20 |
| f3 | SecurityWeek: Zenity Raises $125 Million in Series C Funding | press | 2026-10-20 |
| f4 | AI Defense Matrix Catalog entry | other | 2026-06-13 |
| f5 | AI Defense Matrix Catalog mapping | other | 2026-10-20 |
| f6 | Zenity: Microsoft Power Platform use-case page | official | 2026-10-20 |
This site is an experimental research aid created by Zeltser Security Corp. All its data gathering and analysis was performed autonomously without human review, and it can contain errors of fact, interpretation, and judgment that a human reviewer might catch.
The analyses are statements of opinion, not statements of fact. Machine analysis produced the scores, summaries, and matrix placements by weighing the public sources each page cites, and reasonable people can weigh the same sources differently. Where a page states a fact, it cites the public source and the date it was checked, and the statement is only as accurate as that source. Unless a profile expressly says otherwise, the analysis involves no hands-on testing and no independent validation of any company's products or services.
Nothing here is professional, security, legal, financial, investment, or purchasing advice, and nothing here is a recommendation to invest in, do business with, or avoid any company. Inclusion of a company is not an endorsement, and absence of a company is not a judgment about it. Reading this site creates no advisory or client relationship. Verify any detail you plan to act on against the vendor's current materials.
The content is provided "as is" and "as available," with all warranties disclaimed, express or implied, including merchantability, fitness for a particular purpose, accuracy, and non-infringement. No entry is warranted to be complete, current, or correct. Companies change, vendors update their claims, sources can be wrong, and automated analysis can misread them.
To the fullest extent permitted by law, the operator, Zeltser Security Corp, is not liable for any damages that arise from using this site or relying on its content, including direct, indirect, incidental, special, and consequential damages and lost profits, even if advised that such damages were possible. If you are dissatisfied with the site or disagree with these terms, your remedy is to stop using it.
Entries link to vendor pages, press coverage, and other external sites that Zeltser Security Corp does not control and is not responsible for. A link is not an affiliation with the destination or an endorsement of it. Product and company names and trademarks are the property of their owners, used here nominatively to identify the companies described. Short quotations from cited sources appear for identification and commentary.
Use, quotation, automated retrieval, and redistribution of the content are governed by the Terms of Use at cybercompanyprofiles.com/terms, which permit personal and internal business use with attribution and prohibit republication and resale.