Security & Trust

AdeHQ is built for work. That means trust, isolation, permissions, auditability, and human approval matter from day one.

Workspace data isolation

Each workspace is treated as a separate tenant. Workspace data is scoped by workspace membership and server-side authorisation checks.

Authentication

AdeHQ uses secure authentication through trusted infrastructure providers. Users are responsible for protecting account credentials and using strong passwords.

Role-based access

Workspace owners and administrators can manage users, AI employees, roles, permissions, billing, and workspace settings.

AI employee permissions

AI employees are designed to operate with controlled access. Tools and integrations should be granted deliberately, with read/write/admin levels where supported.

Approval gates

Sensitive AI actions should require human approval. This includes sending external communications, publishing content, making purchases, deleting data, or changing production systems.

Work logs and auditability

AdeHQ records meaningful AI work through work logs, usage records, approvals, memory, artifacts, and runtime metadata. This helps users understand what happened, who did it, and what context was used.

AI provider routing

AdeHQ may use multiple AI providers and gateways. Requests are routed based on reliability, capability, cost, and configuration. Provider details may change over time.

Data retention

Workspace content is retained while needed to provide the service unless deleted or subject to legal, backup, or operational requirements. See our Privacy Policy for more detail.

Responsible disclosure

If you believe you have found a security issue, contact support@adehq.com. Please include enough detail for us to reproduce and investigate the issue.