Security and compliance

How is my content kept safe?

Content is encrypted at rest and in transit, backed up automatically, and every change is written to an audit log you can read. Access is governed by roles that reach down to the individual field, so who may edit, who has to approve, and who is allowed to publish are three separate decisions rather than one permission.

The same rules cover the AI agent. It holds a role, it works inside that role, and its runs appear in the audit log beside everyone else.

Three layers

Encryption at rest and in transit protects the content from the network and the disk. Automatic backups protect it from deletion and from a bad import. The audit log protects it from the hardest case, which is a change nobody will admit to making.

Access is the layer incidents come through

Most content incidents are not attacks. They are somebody publishing a draft they meant to save, or editing the wrong locale, or a contractor who still had access in March.

That is why roles reach down to the individual field and why publishing is a separate right from editing. Somebody can be trusted to write and not yet trusted to put it on the internet, and a permission model that cannot express that pushes teams into sharing one login.

The agent is inside the same fence

The AI agent holds a role, works within it, and its runs appear in the audit log beside everyone else. It cannot reach a field its role cannot reach, and it cannot publish at all.

An agent outside the permission model would make every other control on this page decorative.

Also in security and compliance

All of security and compliance All questions

Still have a question left

Thirty minutes with the product team, and a straight answer.