What our compliance checks do on every outbound call

An AI agent that can place calls needs rules that it cannot skip. Larynth runs the same six checks before every outbound call. If one of them fails, the call is not placed, and the reason is logged where you can see it.

The six checks

1. Verified business. Before an account can place outbound calls, the business behind it is verified. An account that is not verified can only receive calls.
2. Consent on file. The contact needs a consent record: how and when they opted in, and what they agreed to. With no record, there is no call.
3. Do-Not-Call scrub. The number is checked against your own Do-Not-Call list and the Larynth opt-out list, which keeps every request to stop calling.
4. Calling hours. Calling hours are checked for the contact’s state and local time. Outside the allowed window, the call waits.
5. AI disclosure. The agent says it is an AI assistant at the start of the call and names the business it calls for.
6. Encrypted recording. Callers hear a recording notice. Recordings are encrypted and stay in your region, and you choose how long they are kept.

Outbound calls to emergency numbers (911) are also blocked.

Why they run in the engine

These checks are part of the call engine, not a setting on a form. An outbound call goes through them whether it was started from the dashboard or from the API, so a mistake in a campaign cannot place a call that should have been held back.

What stays with you

The checks help you follow the rules. They do not make you compliant on their own, and they are not legal advice. You are the caller, and you are responsible for following the laws that apply to your calls. In practice that means you:

• bring your own Do-Not-Call registrations for outbound campaigns,
• only call people who have given the consent your calls need,
• tell your own staff that calls are recorded, and
• read the terms and acceptable use policy for your situation.

The full detail is on the security and compliance page.