Living DisclosuresEnterprise
Publish a machine-readable, self-updating AI disclosure at the point a governed use case is used, covering who built it, what model answered, and what controls apply.
Living Disclosures (sometimes called point-of-use disclosures) publish what an AI system is and how it's governed at the moment someone actually uses it, rather than relying on a policy document written once and left to go stale. Once a use case reaches active monitoring, the platform can attach a disclosure to every response its agent(s) produce, and generates a public page anyone can open to see the same information without an Airia account.
Where a disclosure appears
A disclosure travels with the traffic it describes, and links out to a page anyone can read.
API responses
Airia agent executions attach disclosure details to the response, plus a link pointing to the public page.
Slack messages
Agents deployed through the AI Gateway that post to Slack get a compact governance footer appended to their message.
A public page
Every active disclosure has a shareable link anyone can open without signing in. It's built for the customer, auditor, or regulator reading it, not the Airia user configuring it.
How a disclosure is built
Two things combine to decide what a given disclosure actually shows:
Disclosure Templates
The reusable policy: which attributes (model details, guardrails, risk classification, and more) your organization is willing to disclose, scoped by audience.
A use case's disclosure
One per use case. Inherits from the applicable template by default, with the option to override individual attributes.
A disclosure only starts publishing once its use case reaches the Monitor stage, even if it was created and activated earlier. See Stage Transitions for the full lifecycle.
An agent that serves more than one use case gets a single merged disclosure. For each attribute, the most restrictive setting across its use cases wins, so no use case can end up disclosing more than it chose to.
Why it matters
Living Disclosures are Airia's control for AI transparency: the requirement that people be able to tell what an AI system is doing and why. Publishing agent identity, model details, guardrail coverage, and risk classification at the moment of use is how the platform satisfies that requirement, rather than relying on a static document that goes stale the moment a model or control changes. See Risks for how disclosures fit into a use case's overall risk and treatment picture.
Permissions
Viewing and managing disclosures are controlled by the Disclosures governance permission. See the permissions reference for details.
Create a Disclosure
Attach a Living Disclosure to a use case and configure what it publishes.