JumpCloud manages identities, devices, and agent access. Willow governs employee AI tools with MCP discovery, approved capabilities, device policies, and runtime approvals.
A 600-person company runs on JumpCloud for directory, SSO and device management. Engineering adopts Claude Code and Cursor, and within a month laptops carry MCP servers nobody approved and skills copied from public repositories. IT can see which apps people sign in to. Security needs more: what those agents are configured to do, which tools each team may call, and a record of what each call contained.
Willow was built for that layer. It discovers local MCP servers, skills and plugins, applies allow, warn and block rules on the device, and gives employees an approved catalog of servers, toolkits and skills through company groups. Runtime guards inspect tool calls, publication checks review what the organization distributes, and approval before selected tool calls keeps consequential actions under human review. Willow runs on top of JumpCloud: it uses JumpCloud for SSO and SCIM, deploys its Scan Agent through JumpCloud and reads device ownership from JumpCloud.
JumpCloud approaches AI from the directory. Its Agentic IAM, announced in April 2026 as an extension of its identity and device management platform, registers agents as identities tied to human owners. AI & SaaS Management discovers AI usage through the JumpCloud Go browser extension and the JumpCloud Agent, and the AI Gateway acts as a centralized MCP control plane. JumpCloud is rolling these capabilities out through 2026.
Both products now touch MCP access and shadow AI. The difference is depth. JumpCloud extends identity, group membership and device trust to AI clients and agents. Willow governs the AI capabilities themselves: the configuration on each device, the tools exposed on each server, the content of each call and the review before sensitive ones run.
The audit row deserves attention. JumpCloud documents that AI Gateway events don't include user prompts, tool arguments, tool responses, or credentials. Willow's request logs include a details view with the full event payload, and its runtime guards evaluate tool arguments and responses before they pass. When an incident involves what an agent sent or received, that difference decides whether security can answer.
An agent's behavior is set long before it signs in. An MCP server added to Cursor, a skill installed in Claude Code or a CLAUDE.md file with standing instructions all change what it will do. Willow's Scan Agent inventories those capabilities, and JumpCloud customers can roll it out with JumpCloud device management to macOS and Windows and attribute each device to its owner with a read-only service account.
Findings become allow, warn and block rules scoped to everyone, specific groups, users or devices. A block rewrites the MCP entry to a local Willow endpoint or renames the skill file, and the agent restores the original when the rule is removed. Willow can also push approved agent configuration to Claude Code, Claude Desktop, Codex, Cursor and GitHub Copilot, currently in beta. Enforcement happens on the device, not only in the browser.
Granting a team access to an MCP server often grants every tool it exposes. Willow supports tool-level availability and human approval before selected tools run. A team can keep read operations open, disable write tools, and require confirmation for a sensitive invocation. Approval applies to every call to a tool marked for it.
Willow's runtime guards inspect a tool call's content before it reaches the MCP server and the server's response before it returns to the client, and can block or mask flagged content. Guard Hooks bring the same guards into Claude Code, Cursor and Codex for prompts and file-tool calls; they fail open on a timeout or outage. Prompt Guard applies them to browser AI chats in Chrome and Edge, in beta with interception off by default.
Willow groups decide which MCP servers, toolkits and skills each employee can use, and SCIM from JumpCloud keeps those groups current. Organizational skills sync from GitHub and bundle into plugins, and build-time guards block a toolkit or plugin from publishing when its content violates policy. Employees get approved tools; security reviews them before they spread.
Machine users carry accountable human owners and access limited to the MCP servers assigned to their groups, and the Vault keeps connector secrets out of configurations. Radar scores AI security risk across identity, shadow AI, data exposure, agents and monitoring. Willow deploys as SaaS, hybrid or full on-premises, with no runtime dependency on Willow SaaS in the on-premises model.
JumpCloud fits organizations that want one vendor for directory, SSO, MFA and device management, and want AI access handled in the same console:
These are strong directory and device controls. They work best alongside a layer that governs what each AI tool is configured to do and what each call contains.
As of 28 October 2026, Willow lists a $0 Free plan and Enterprise pricing per human seat, billed annually, with volume tiers. Enterprise includes shadow AI discovery across the fleet, runtime guardrails, browser prompt guards and SCIM, and Willow can be purchased through AWS Marketplace. Willow deploys as SaaS, hybrid or full on-premises.
JumpCloud lists self-serve packages from $9 per user per month billed annually for Device Management, with Platform tiers priced through sales and a 30-day free trial. Agentic IAM Lifecycle Management appears in its pricing table without a listed price, and the AI Gateway documentation notes that functionality and pricing may change before general availability. JumpCloud runs as a cloud service with US, EU and India regions.
For most JumpCloud customers, yes. JumpCloud stays the system of record for people, groups and devices. Willow inherits those through SSO, SCIM and device attribution, and deploys its Scan Agent through JumpCloud. Willow then governs the AI layer: local MCP servers and skills, the approved catalog, tool-level availability, content inspection, approvals and audit.
Where both reach MCP access, route employee AI clients through one gateway to avoid duplicate policy. Willow's per-tool controls and payload-level audit make it the natural owner of that path for security-sensitive teams.
JumpCloud gives IT a single place for identities and devices, now including agents. Willow turns that foundation into governed AI adoption: what runs on each device, which tools each team gets, what every call contains and which calls need review. To see Willow working with your JumpCloud directory, book a demo or start on the Free plan.
Yes. Willow uses JumpCloud as an SSO provider through a custom OIDC app, supports SCIM provisioning from JumpCloud, deploys its Scan Agent through JumpCloud device management, and reads device ownership from JumpCloud with a read-only service account.
JumpCloud's Agentic IAM feature brief describes discovering MCP configurations on devices, and its April 2026 announcement lists AI discovery among capabilities rolling out through 2026. Willow's Scan Agent discovers MCP configurations, skills, plugins and instruction files today and enforces policy rules on the device.
JumpCloud documents that AI Gateway events don't include user prompts, tool arguments, tool responses or credentials. Willow request logs include the full event payload for requests it handles, filterable by user, agent, server, tool and guard action.
As of 28 October 2026, JumpCloud lists the AI Gateway as a public preview and notes that functionality and pricing may change before general availability.
Yes. Full on-premises deployment runs every Willow component in the customer's Kubernetes cluster, with no dependency on Willow SaaS at runtime.
Give them a Basecamp. Go from AI chaos to AI work, in minutes.
