Permissions Explorer

Control who can access knowledge, capabilities and actions.

AI Employee OS separates organizational access, Knowledge permissions, Agent capabilities and Workflow approvals into understandable layers.

Start free trial Explore Enterprise

Eight permission layers

Select a layer to understand its responsibility.

Permission layer

Company

The top-level boundary for ownership, billing, governance and organization-wide policy.

Typical controls

Company administrators

Global policy

Data ownership

Access is cumulative and explicit: a connected system does not automatically grant an AI Employee, Agent or Workflow permission to use it.

Control principles

Permissions should be understandable before they are powerful.

Least privilege

People and AI Employees receive only the access required for their responsibility.

Explicit assignment

Knowledge, Agents and Workflows are assigned deliberately instead of inherited without review.

Visible boundaries

Administrators can understand which layer grants access and where approval is required.

Human control

Sensitive actions can require named reviewers before a Workflow continues.

Knowledge

Control approved sources and Collections.

Open Knowledge Center

AI Employees

Assign Knowledge, Agents and audiences.

Open Builder

Workflows

Place approval gates before controlled actions.

Open Workflow Builder

Your organization defines the boundaries.

AI Employee OS applies those boundaries across Knowledge, AI Employees, Agents and Workflows.

Explore Enterprise control