Venice is modern privileged access management (PAM) for the agentic era. It replaces standing privilege with just-in-time access for people, non-human identities (NHIs) and AI agents across cloud, on-premises and SaaS environments. The company emerged from stealth in February 2026, operates from Tel Aviv and New York, and is led by co-founder and CEO Rotem Lurie alongside co-founder and CTO Or Vaknin. Where vault-based PAM adds a checkout step but leaves the privilege behind it standing, Venice grants privileged access at the moment of a request, scopes it to the task, and revokes it when the task ends, using an agentless deployment model. In June 2026 the company extended this model to AI agents with the MCP Gateway, part of its Tailored Access capability. This overview explains how the platform works and what it covers, based on Venice’s public product information and announcements.
At A Glance
| Vendor | Venice, with operations in Tel Aviv and New York |
|---|---|
| Category | Modern privileged access management for the agentic era; zero standing privilege, just-in-time access |
| Identities covered | People, non-human identities (service accounts and workloads) and AI agents |
| Deployment | Agentless, with one outbound-only connector per network zone; covers cloud, on-premises and SaaS environments |
| Access model | Just-in-time, task-scoped privilege granted per request and revoked automatically |
| Agentic AI | MCP Gateway, part of Venice Tailored Access: a runtime control point for AI agents |
| Funding | $33 million in total, including an $8 million seed round led by Index Ventures and a $25 million Series A led by IVP |
What The Platform Does
Venice was built as one platform with one policy engine for every identity type, so people, NHIs and AI agents share the same policy, risk scoring and audit trail. Its capabilities fall into five areas.
Zero Standing Privilege And Discovery
Venice continuously discovers privileged identities and standing access across environments, flags what is risky, and replaces unnecessary standing privilege with just-in-time access. A Zero Standing Privilege dashboard shows where standing access exists, the risk it introduces, and the organisation’s progress towards eliminating it, with AI-generated insights and KPIs. People, NHIs and agents appear in one view.
Just-In-Time Access
Users request, approve and connect to systems across cloud, on-premises and SaaS environments from a single browser-based interface, or from Slack, Microsoft Teams or an MCP client. Access is granted at the moment of the request, scoped to the task and removed when the task ends. Venice states that adding a server, database or cloud resource takes minutes through one-click onboarding, converting static privileges into dynamic, policy-driven access.
Context-Aware Access
Venice evaluates each request against a live policy assembled at the moment of the request. The company describes four inputs:
- Identity risk and posture, including account risk score, historical access patterns, group memberships and transitive permission paths
- Environment and device context, including device posture, network segment, cloud provider and environment classification
- Session intelligence, where active sessions are monitored for anomalous behaviour and a flagged session increases scrutiny of the identity
- One policy engine, which approves low-risk routine requests automatically and routes higher-risk requests to a human reviewer with context already assembled
Legacy PAM Migration
Venice scans existing vaults, shared accounts and standing permissions to map how privilege has been granted, then uses an AI-assisted migration tool to translate legacy permissions into policy-driven just-in-time access. Shared accounts can be kept during the transition, with credential rotation and usage monitoring. Venice is designed to run alongside an existing PAM tool and roll out environment by environment; published customer stories describe migrations from CyberArk and Delinea.
Audit And Session Accountability
Venice links every request to a decision, every decision to an access grant and every grant to a recorded session, each attributed to a named identity. A searchable Audit Center covers requests, approvals, access events, policy changes and session recordings, which Venice says are enhanced with AI. The platform generates scheduled reports for quarterly certifications, SOC 2 evidence and cyber insurance renewals, and streams enriched, structured access events to SIEM platforms.
How The Architecture Works
Venice’s product pages and announcements describe the following model.
| Component | What it does |
|---|---|
| Agentless deployment | Venice requires no agents on target systems or user machines, and no VPN. One outbound-only connector is deployed per network zone. Infrastructure is onboarded centrally and existing standing privileges are converted into just-in-time policies. |
| One policy engine | Every request from every identity is decided by the same policy engine, which weighs who is asking, the context and the task. For people and service accounts, this is where requests are decided and sessions recorded. AI agents reach the same engine through the MCP Gateway. |
| Elevation on the target system | When a request is allowed, Venice provisions the specific short-lived access the task needs on the target system and rolls it back when the work is done. Where a credential is needed, Venice vaults and rotates it and injects it into the session, so the user or agent never handles the secret. |
| Shared risk engine and audit trail | Every identity type is scored by the same risk engine and written to the same audit trail, so people, NHI and agent activity is recorded in one place. |
Agentic AI Capabilities
Venice treats AI agents as one more identity type, governed by the same just-in-time model it applies to people and service accounts.
MCP Gateway
Announced in June 2026 as part of Venice’s Tailored Access capability, the MCP Gateway sits between agents and the resources they use. To an agent, Venice appears as an MCP server; in practice it is the control point in front of every system the agent can reach. Venice describes the Tailored Access capability as combining intent and ownership detection, contextual controls and runtime authorisation, credential management, just-in-time access and full audit.
Agent Discovery And Ownership
Venice connects to the platforms where agents run and discovers agents along with their owners. Each agent is listed beside humans and service accounts on the identity page of the person it answers to. During onboarding, the agent’s job is mapped to a scoped slice of its owner’s access, limited to what the task needs.
Runtime Authorisation And Drift Detection
Agents must request access against a declared task. Each call is evaluated against that task, the session’s running history, the human owner and risk signals such as unusual volume or a step up in sensitivity. A call that fits proceeds; one that drifts is escalated for approval, or the session is ended and its access revoked before the call completes. Venice states that agents never hold credentials of their own, so there is no secret in the agent’s context to leak.
Agent Session Audit
Venice records each agent request, access grant and session action, and traces actions back through any chain of agents to the accountable human and task. Its automated session analysis shows where an agent stayed within scope, the exact point any call drifted, and the access that was elevated and rolled back.
Integrations
| Area | Documented integrations |
|---|---|
| Target environments | Servers, databases and cloud resources across cloud, on-premises and SaaS, including identity provider roles and on-premises directories |
| Access requests | Browser interface, Slack, Microsoft Teams and MCP clients |
| AI agents | MCP, through the Venice MCP Gateway, with Venice acting as an MCP server in front of agent tool calls |
| Legacy PAM | Analysis and migration of existing vaults and shared accounts; published customer stories describe migrations from CyberArk and Delinea |
| Security operations | Streaming of structured access events to SIEM platforms |
| Channel partners | Consortium, Crush Security, Cualified, GuidePoint Security, Omada, Optiv and Trace3 |
Use Cases And Industries
Venice publishes use cases for securing AI agents, achieving zero standing privilege, just-in-time access, replacing legacy PAM, context-aware access and audit sign-off. Its customer stories describe a Fortune 1000 automotive company with no prior PAM that reached production in 48 hours and had every user identity managed by Venice within five business days; a global industrial manufacturer that migrated 2,000 privileged accounts and more than 800 servers from Delinea in three weeks; and a global entertainment company that replaced CyberArk and removed 99% of standing privileges on its SOX servers. Venice displays SOC 2 and GDPR compliance badges on its website.