# Security (https://riska.io/security)

_Security_

## Data governance for the AI age.

When agents and automations act on your book, audit logs, security and attribution for every change are requirements. Riska builds them in from the start.

## Your book stays yours.

Governance is part of the platform, not a setting added afterwards.

- **A deployment of your own**: Every customer runs in a dedicated deployment, set up around their own governance needs.
- **Data that is never shared**: All of your data stays in your deployment. It is never pooled or shared with anyone else.
- **Your access, your rules**: Single sign-on through your identity provider, and roles that set what each person can see and change.
- **Permissions that bind AI**: Ask Riska and every agent work with the access of the person they act for, never more.
- **Every change attributed**: An audit trail shows who changed what, whether a person, an agent or an integration, and every AI decision is kept on the record.
- **No training without permission**: We don't train models on your data unless you explicitly allow it, for use cases you want in your own workspace.

Found a security issue? Write to [security@riska.io](security@riska.io) and see our [Vulnerability Disclosure Policy](/disclosure-policy).

## Want the details?

Book a demo and we'll walk your team through how Riska is deployed and governed for you.