Audact Inside
Audact Inside is the part of a governed interaction that faces outward: the disclosure that is spoken, the mark beside a chat message, the attribution in a white-label dashboard, and the receipt each of them points at.
It is not an extra product and there is nothing to switch on for the evidence itself. Every governed interaction leaves a receipt whether or not anyone ever looks at it. Inside is what makes it findable.
Where it surfaces
Voice, chat and dashboard show the signal differently because the medium differs. What they point at is the same object.
Voice
In controlled testing
At the position the policy for that country requires (Germany active · Netherlands in progress · Belgium, Austria queued), the agent states that the caller is speaking with an AI system. Where the customer enables it, the flow also names where the record of the call can be requested. The receipt records the disclosure step: its wording hash, the jurisdiction and its position in the call.
Chat
In development
A small mark sits next to AI-generated messages. Opening it shows the receipt reference for that exchange and lets the reader copy it, so a screenshot is not the only thing they walk away with.
Dashboard
In development
Inside an agency deployment, the end client sees a discreet attribution. It is the agency's decision whether to show it; the evidence is produced either way.
What can be checked
This is the part that makes the signal worth showing: three values on the receipt that anyone holding it can recompute, without a key and without an account. It is the same answer as on the proof page.
Without a key, without an account
The authentication code is HMAC-SHA256 under a symmetric key held by Audact; whoever can check it can also produce it. A receipt is a faithful record of what the system received, decided and released. That is what it establishes, not whether the world matched it.
The path
A call is answered or placed, a chat turn is generated, a dashboard is opened.
The disclosure is spoken, the mark appears beside the message, or the attribution shows in the dashboard.
The receipt reference identifies one governed interaction. Today it is requested through the customer whose agent placed the call.
The receipt explorer walks a chain field by field with the sample files next to it. No account, no API key.
Walk a chain field by field.
Field by field, and exactly what a receipt is a record of.
What the signal says
Most trust marks on the web assert that someone examined something. Audact Inside asserts something narrower and more useful: that this specific interaction produced an artefact, and that three values in that artefact can be recomputed by anyone who holds it — with a fourth, Merkle inclusion, checked against the published root with the inclusion proof from the call's Article 50 report. No authority has assessed the deployment behind it, and the mark does not claim one has.
That is why it is per interaction rather than per company. A yearly assessment says something about an organisation; a receipt says something about the call your client actually received.
Sub-processors and hosting.
Brand integration
For brand teams and agency customers deciding how the disclosure sounds, where the mark sits and whether the attribution shows. It is a conversation with a person, not a settings page.
Placement options and white-label questions.
How a white-label deployment is set up.
