Does an AI receptionist have to say it's AI?

The federal rules most people have read about — the TCPA, and the FCC's 2024 ruling on AI voices — govern calls a business places, not calls a customer places to you. No federal rule requires you to tell callers your receptionist is an AI. Vona tells them anyway, on every call, in every language.

The short answer

If a customer calls your business and an AI answers, that is legal. The Telephone Consumer Protection Act (TCPA), which most people think of when they hear "AI calling regulations," governs outbound calls — when you call customers. It does not restrict inbound calls, where the customer calls you.

There is no federal disclosure mandate for inbound AI receptionists. A handful of states have drafted AI transparency laws, but they target specific scenarios — companion chatbots, automated hiring systems, deepfakes — and most do not reach a business answering its own phone.

If your AI receptionist records calls, the law depends on your state's wiretap statute, which varies by jurisdiction. Some states allow recording with one-party consent; others require all-party consent.

Inbound versus outbound

The US legal system draws a sharp line between calls coming in and calls going out. Understanding this distinction is the key to understanding what is and is not regulated.

Inbound calls: a customer dials your business number and an AI picks up. This is not the conduct the TCPA's robocall restrictions were written for, and no federal rule requires you to announce that the voice answering is an AI.

Outbound calls: Your business calls customers using an AI voice. Governed by the TCPA. Requires prior express written consent from the customer, proper caller identification, and an opt-out mechanism. The FCC confirmed in February 2024 that AI-generated voices fall under these rules.

Vona is an inbound receptionist. She sits on your business line and answers when customers call. This is the legal space where she operates.

What the FCC ruling actually says

On February 8, 2024, the FCC unanimously ruled that AI-generated voices are covered by the TCPA's restrictions on artificial or prerecorded voice calls. This does not make AI calls illegal. It means they must comply with existing consent and identification rules — and it applies to outbound calls only.

The ruling was prompted by bad-faith actors using voice cloning to impersonate candidates and public figures in robocalls. The FCC's move was to give state attorneys general tools to prosecute those scams. It does not reach a business answering its own phone line.

State-level AI laws

A few states have begun drafting AI transparency laws. California, for instance, passed a "companion chatbot" law in 2025, effective January 2026, that requires disclosure when an AI system simulates a human-like relationship and the user could reasonably be misled. The law targets chatbots designed to create emotional bonds, not professional business receptionists.

Where a rule does apply, what it asks for is usually light — a line of disclosure at the start of the call. But this area is changing quickly, and whether a given law reaches your business is exactly the kind of question to put to your own lawyer rather than to a vendor's website.

Call recording and consent

If your AI receptionist records calls — whether for training, compliance, or transcription — the law depends on where the call is being made from.

One-party consent (38 states plus DC): The business can record without informing the caller. Applies to: AK, AL, AR, AZ, CO, CT, GA, HI, IN, IA, KS, KY, LA, MI, MN, MO, NE, NJ, NM, NC, ND, OH, OK, OR, RI, SC, SD, TN, TX, UT, VT, VA, WI, WY, and DC.

All-party consent (11 states): Every participant must consent before recording begins. Applies to: CA, DE, FL, IL, MD, MA, MT, NV, NH, PA, WA.

When a call crosses state lines, the stricter law applies. If either party is in an all-party consent state, all parties must consent.

Disclaimer

Frequently asked questions

• Is it legal to have an AI answer inbound business calls?

  The federal rules people worry about — the TCPA, and the FCC's 2024 ruling on AI voices — are written for calls a business places, not calls a customer places to you. No federal statute bars a business from using AI to answer its own line. State law varies, so check yours.

• Does an AI receptionist have to tell callers it's an AI?

  No federal rule requires it for inbound business calls. Several states have passed AI transparency laws aimed at various situations, and more are being written — whether one covers you is a question for your own counsel. Vona discloses regardless, on every call, in whichever language the call is in.

• What's the difference between inbound and outbound AI calls?

  Inbound is a customer dialling your business and an AI picking up. Outbound is your business placing calls with an AI voice — that is the TCPA's territory, requiring prior express consent, identification and an opt-out. US law treats the two very differently, and most coverage you have read concerns outbound.

• If an AI receptionist records calls, does it need consent?

  It depends on your state's wiretap law. Thirty-eight states plus DC allow recording with one-party consent — the business can record without telling the caller. Eleven states (California, Delaware, Florida, Illinois, Maryland, Massachusetts, Montana, Nevada, New Hampshire, Pennsylvania, Washington) require all-party consent — every participant must agree. When calls cross state lines, the stricter law applies.

• Are there state-level rules about AI in customer service?

  Several states have begun legislating AI transparency, usually aimed at particular uses — California's companion-chatbot law, effective January 2026, was written for systems that simulate a human-like relationship. The field is moving quickly. Put the question to counsel in your own state rather than relying on a summary like this one.

• What does Vona do?

  Vona answers on the first ring, 24/7, in English and Spanish. She qualifies callers, books appointments, and sends the owner a text summary. She always tells the caller that she is an AI, in whichever language the call is in. She never pretends to be human. Call the demo line to hear her: +1 (561) 566-1668.

Don't take a website's word for it

Ask her yourself. If you would rather we set her up on your phones, request a demo.