There's a temptation in the AI phone business that nobody says out loud: make the voice so human that callers never ask. Some vendors treat that as the goal. We think it's a mistake — legally, commercially, and on plain principle. Here's where the rules actually stand, where they're headed, and why the honest configuration wins anyway.
Context for what follows: the receptionist is the part of our system you can call and test — (626) 365-4946 — but it's one component. The build behind it runs capture, booking, follow-up, reviews, reactivation, and reporting as a single system. Keep that frame; this piece zooms in on one gear.
The rules are moving in one direction
The regulatory picture for AI calls is a patchwork, but the direction of travel is not ambiguous. Major platforms and telecom providers increasingly require that automated and AI-generated voice systems be identifiable as such. Several states have introduced bills specifically targeting AI disclosure on calls — Illinois SB 317, which would have required AI systems to affirmatively disclose themselves, stalled in committee, but nobody watching the space believes that's the end of the story. Bills like it keep appearing, and each generation gets closer to passing.
Building a phone operation on the assumption that concealment stays legal is building on sand. Building on disclosure means the next law changes nothing for you.
Our stance: admit it when asked, always
Our position is simple and non-negotiable: if a caller asks whether they're talking to an AI, the answer is yes. Every time, in plain language, without deflection. Not "I'm a virtual assistant" delivered in a tone engineered to end the inquiry — an actual answer.
Why? Because the alternative fails on its own terms. A caller who suspects, asks, and gets dodged now distrusts two things: the voice on the phone and the business that put it there. The deception doesn't even buy anything. Callers don't hang up because a competent AI answered; they hang up because nobody answered, or because something felt off. Straight answers are what "not feeling off" is made of.
No voiceprints, period
There's a second, quieter issue in AI telephony: biometric identification. Voice is a biometric. Systems that fingerprint callers by their voice — recognizing them across calls without being told who they are — are collecting biometric identifiers, and in Illinois that runs straight into the Biometric Information Privacy Act (740 ILCS 14), one of the strictest biometric laws in the country, with a private right of action attached. BIPA lawsuits have cost companies enormous sums.
Our answer is not "we handle voiceprints carefully." It's that we don't voiceprint-identify callers at all. A caller is who they say they are, the way it works when a human answers the phone. The cleanest compliance posture for a biometric risk is not having the biometric.
Recording off by default
Call recording is the same story. Consent rules for recording vary by state — some require all parties to consent — and a national caller base means you're effectively subject to the strictest rule that applies to any caller. So our default is that recording is off. If a business has a legitimate need and a proper consent flow, that's a deliberate decision made with eyes open — not a checkbox that was silently on from day one. Transcripts of what the AI needs to do its job (book the appointment, log the request) don't require keeping audio of a person's voice forever.
Honesty converts better than tricks
Here's the part that surprises people: none of this costs conversions. In our experience it does the opposite. A caller at 9pm who gets an immediate answer, real available times, a clear read-back of their appointment, and a straight "yes, I'm an AI — want me to transfer you to a person?" walks away with something no evasive system produces: the sense that this business is competent and has nothing to hide.
Trust is the actual product. The appointment is just the receipt. If you want to hear what a fully disclosed, no-voiceprint, recording-off configuration sounds like on a live call, book a demo and ask it anything — including whether it's a robot.
Hear what honest AI sounds like on a real call — book a demo or dial (626) 365-4946, and yes, an AI answers; that's the point.