Updated 26 Oct 2026

AI disclosure policy

How we label AI-made content and what we will never generate.

Draft Draft for counsel review. Not yet in force; the final version will be published here with its effective date.

What is AI-made

Drafts in Kirtis are written with AI models from facts you confirmed. Media made with an AI host or an AI voice is marked AI-made in the app and in the published post, with the network's AI flag where it has one.

Hosts and personas

An AI host needs two separate approvals before it appears: one for how it looks and sounds, one for how it is used. A host never presents itself as a real customer or a real employee who does not exist.

What we will not generate

  • Reviews or testimonials, by anyone, real or synthetic.
  • Content that imitates a real person without their written consent.
  • Numbers about your company that you have not confirmed.
  • Clones of another brand's ads, emails or posts.

Your part

You approve what is published under your name. If you remove an AI-made label where the law requires one, the responsibility is yours, and we may pause publishing for the workspace.