The three limits: do not deceive the caller, do not impersonate a named person, and follow your state's call-recording consent rules. This post covers the short answer, the nuance behind it, and a safe-practice checklist. It is general guidance, not legal advice, so confirm your own state before you launch.
The short answer, then the nuance
Using AI to answer inbound calls and book appointments sits on solid ground. The caller dialed you. The AI helps them get what they wanted, which is an answer and a scheduled visit. Nothing in that transaction runs afoul of federal law.
The nuance lives in three areas: what the FTC treats as deception, how call-recording consent works in your state, and a small set of bot-disclosure laws aimed at other uses. Handle those three and you are clear. Ignore them and the risk is not the AI itself, it is how you configure it.
What the FTC cares about: deception, not the tool
The FTC polices unfair and deceptive practices. Its concern is not that a business uses AI. Its concern is whether a business lies to a consumer. An AI that answers honestly and helps a caller book a repair is not deceptive.
The line you do not cross is impersonation. Do not build an AI that claims to be a specific human employee, invents a fake name and backstory, or denies being AI when a caller sincerely asks. Give the assistant a clear name and a clear job, let it help, and answer honestly if someone asks whether they reached a computer. Honesty there builds more trust than a dodge, and most callers never ask, because the booked appointment is the thing they wanted.
Call recording and two-party consent
Recording is a separate issue from AI, and it trips up more businesses than the AI question does. Some states require only one party to consent to a recorded call. Others require every party to consent. If you save call audio for training or quality review, the recording rules apply whether a person or an AI is on your end of the line.
The fix is simple. If your state requires all-party consent, add a short recording notice at the start of the call, the same disclosure a call center uses. Check your own state's rule before you turn on recording, and when you serve callers across state lines, follow the stricter standard.
Bot-disclosure laws aim elsewhere
A handful of states have bot-disclosure laws, and they scare owners who read the headline and skip the scope. Those laws target specific conduct: using a bot to sell something or to influence a vote without disclosing that it is a bot. California's bot law, the most cited one, covers commercial sales and political messaging aimed at influencing a purchase or an election.
Answering an inbound service call and booking a repair is neither. The caller came to you for help, not the reverse. Still, laws change and vary by state, so treat clear naming and honest answers as the standard everywhere. It keeps you inside every version of these rules and costs you nothing.
Safe-practice checklist
Five rules keep you clear. Give the AI a clear name and a clear role instead of a fake human identity. Never let it deny being AI when a caller sincerely asks. Configure it to answer from your real business facts so it does not invent claims. Add a recording notice if your state requires all-party consent. And confirm your own state's rules with your attorney before launch, because this is general guidance, not legal advice.
Follow those and the legal question settles into the background, where it belongs. For the practical side of what callers actually hear and whether they can tell, read will customers know they're talking to a bot. To see how an AI front desk fits a service business, visit the contractor page.
Common Questions
Do I have to tell every caller they reached an AI?
No general law requires an unprompted disclosure at the start of a booking call. The rule is narrower: the AI must not deny being AI when a caller sincerely asks, and it must not impersonate a specific human. If someone asks outright, an honest answer is the right move and it builds trust.
Does the FTC ban using AI on business calls?
No. The FTC targets deception, not the tool. An AI that answers honestly and helps a caller book a service is fine. The problem it would flag is an AI that impersonates a named employee or lies about what it is.
What about recording the calls?
Recording is governed by your state's consent law, separate from the AI question. Some states need one party's consent, others need all parties. If yours requires all-party consent, add a short recording notice at the start of the call, and follow the stricter standard when you serve callers in multiple states.
Do state bot-disclosure laws apply to my answering service?
They aim at bots used for commercial sales or political influence without disclosure, not at answering an inbound call and booking a repair. Even so, laws vary and change, so clear naming and honest answers keep you inside every version of the rule. Confirm your own state with your attorney before launch.
the right way.
We will walk through the setup and the disclosure choices for your state and your trade.
Book a Free 30-Minute Call