Honest about being an AI
- If a caller asks whether they are talking to a robot, the AI voice agent says it is an AI
- A caller can ask for a person at any time
- Our trained human agent can take over the same call
Trust and safety
VoiceBridg tells callers it is an AI when they ask, and a caller can reach a person at any time. Gas and carbon monoxide calls follow a safety script first. Transcripts are encrypted and phone numbers are masked in the interface. We don't hold a SOC 2 report or HIPAA attestation yet, and we say so.
Last updated October 31, 2026
Live call · answered in 1 ring
Sample callTomorrow · 8:00 AM
AC repair · 412 Maple St. · Booked
We don't yet hold a SOC 2 report or a HIPAA attestation. Call-recording and AI-disclosure rules differ by state, and which ones apply to you is for you and your adviser to decide. This is not legal advice. Tell us your requirements before you start and we'll say plainly what we can and can't support today.
VoiceBridg gives you the controls that matter in the call path: AI disclosure, a way to reach a person, safety scripts, recordings with an audit trail, and a kill switch. Which rules apply depends on your state and your business, so check with your own adviser. This is not legal advice.
If a caller asks whether they are talking to a robot, the AI voice agent says it is an AI and offers a person. It never pretends to be human.
Safety comes first. The caller follows a safety script, and the call is logged for your team rather than booked as a routine job.
Yes. The AI kill switch stops AI handling across your workspace, including calls already in progress, and routes callers to people.
Only people you give access to, under role-based permissions. Transcripts are encrypted with AES-256, phone numbers are masked in the interface, and you set how long recordings are kept.
Not yet. If you need either, tell us before you start and we'll say plainly what we can and can't support today.
Looking for the software platform's controls? See VoiceBridg Platform trust and security.