Trust & Objections · informational
Does an AI Receptionist Have to Disclose It Is an AI? (What the Law Actually Says)
Plain-English breakdown of AI receptionist disclosure law — FTC guidance, state rules, and FCC provisions — and why transparency actually books more jobs.
Your competitor’s phone rings at 9 PM on a Tuesday. Nobody picks up. A homeowner with a burst pipe moves on to the next result in Google. That’s a $1,200 job gone — not because of pricing, not because of reviews, but because no one answered.
An AI receptionist solves that problem. But before you set one up, there’s a question your attorney, your customers, or a curious journalist might eventually ask: are you legally required to tell callers they’re speaking to an AI? And what happens if you don’t?
The answer is more nuanced than a yes or no — and the practical upshot may surprise you.
The Federal Picture: FTC Authority and the “Deception” Standard
There is no single federal law titled “AI Voice Agent Disclosure Act.” What exists instead is the Federal Trade Commission’s long-standing authority to pursue businesses for unfair or deceptive acts or practices under Section 5 of the FTC Act.
The FTC has made clear in its AI guidance that using AI to impersonate a human in a commercial context — when that impersonation is material to the transaction — is the kind of deceptive practice it will pursue. The FTC’s 2023 policy statement on commercial surveillance and its 2024 AI guidance both reinforce that consumers have a right to know when they are interacting with a machine, particularly when that machine is influencing a purchasing decision.
What does “material” mean in this context? If the caller would behave differently — provide less personal information, ask different questions, or simply hang up — knowing they were talking to an AI rather than a person, the deception is considered material. Booking a $500 AC repair appointment generally clears that bar.
In practice, the FTC has focused enforcement on outbound robocalls and AI-generated voice cloning scams, not inbound service business lines. But its enforcement authority is real, and small businesses are not exempt. The risk is low today; it is not zero, and it is rising as regulators pay more attention to commercial AI.
State Law: California’s Bot Disclosure Act and What It Actually Covers
California’s BOT Disclosure Act (Business and Professions Code §17940, passed via AB 1950) is the most commonly cited state law in this conversation. It requires disclosure when a bot is used to communicate with a California consumer with the intent to sell a product or service — but its text is specifically aimed at “online platforms” and written/chat interactions.
Phone calls are not squarely within the Act’s plain language as currently written. That said, California’s broader consumer protection law (the Unfair Competition Law, Business and Professions Code §17200) is wide enough to reach deceptive conduct in any medium, including phone. California regulators have shown appetite for broad interpretation.
Other states have passed or are actively considering similar laws. Illinois, Texas, and New York have all had AI disclosure bills introduced in recent legislative sessions. The trend is clearly toward more disclosure requirements, not fewer.
If any of your customers are in California — which is true for most national or multi-state service businesses — the conservative read is to disclose.
The FCC’s Role: Robocalls, TCPA, and Artificial Voice Rules
The Telephone Consumer Protection Act (TCPA) and the FCC’s implementing rules are often confused with AI disclosure requirements. They overlap but are distinct.
The TCPA restricts automated dialing and pre-recorded messages — it’s aimed primarily at outbound marketing calls. If your AI receptionist answers inbound calls from customers who called you, TCPA consent requirements do not apply in the same way they do to outbound campaigns.
However, in 2024 the FCC issued a ruling clarifying that AI-generated voices used in robocalls (including political calls and commercial solicitations) are subject to existing restrictions on “artificial or prerecorded voice” calls. If your AI receptionist is ever used to make outbound calls — follow-up calls, appointment reminders, lead callbacks — those calls must comply with TCPA consent rules, and the AI nature of the voice is relevant to whether the call is considered “artificial voice” under the rules.
For a straightforward inbound service line, the TCPA is not your primary regulatory concern. But it becomes relevant the moment you move toward outbound use.
What “Disclosure” Actually Has to Look Like
Assuming you decide to disclose — which you should, for reasons both legal and practical — what does that mean concretely?
You do not need a long disclaimer. You need one sentence, delivered early in the call:
“Hi, this is an AI assistant for [Business Name] — I can get you scheduled right now or answer any questions.”
That’s it. No legalese. No apology. You state what you are, and you immediately pivot to what you can do. Studies of small-business call handling consistently find that callers care far more about whether their problem gets solved than about whether the voice belongs to a person or a machine. Clarity builds trust faster than ambiguity does.
What you should not do:
- Have the AI claim to be a named human employee (“Hi, this is Sarah at Riverside HVAC”)
- Use a voice and name specifically designed to make callers believe they are speaking to a person
- Deny being an AI if a caller asks directly
That last point is worth underlining. If a caller asks “Am I talking to a robot?” and the system says “No, I’m a real person,” that’s where the legal and ethical problems concentrate. No competent AI receptionist vendor ships that behavior, but verify before you deploy.
The Honest Objection: “Won’t Disclosure Kill the Call?”
This is the concern every service business owner raises, and it deserves a straight answer.
The evidence does not support it. The Real Stats: How Often Do Callers Hang Up on AI Answering Services? covers this in detail, but the pattern holds: callers who are surprised mid-conversation that they’ve been talking to an AI — and who feel tricked — hang up at much higher rates than callers who knew from the start. The deception, once noticed, is the problem. Not the AI.
There is also a selection effect working in your favor. Someone who calls your HVAC number at 11 PM with no air conditioning is not shopping for a human connection. They are shopping for a next-step. An AI that answers, confirms availability, and books a morning appointment has delivered exactly what they needed.
For more on how caller expectations actually work, see Will Customers Actually Talk to an AI Receptionist? The Complete Trust Guide for Local Service Businesses.
When a Human Still Wins — and How to Hand Off Cleanly
Disclosure is also the moment you can set expectations for escalation. If a caller has a complex billing dispute, a situation that requires judgment beyond booking, or simply says “I’d prefer to speak with someone” — a properly built AI handles that gracefully.
The handoff matters as much as the disclosure. How to Transfer a Call from AI to a Human Without Losing the Caller walks through how to do this without dead air, dropped calls, or a frustrated customer who has to explain themselves twice.
A well-designed AI receptionist treats human escalation as a feature, not a failure. Mentioning it during the disclosure phase — “if you’d rather speak with someone directly, just say so” — is both honest and effective.
A Practical Compliance Checklist for Service Businesses
| Area | What to Check |
|---|---|
| Call opening | Does the AI identify itself as AI in the first sentence? |
| Direct question handling | Does the system correctly answer “are you a bot?” |
| Outbound use | Are any outbound AI calls covered by TCPA consent? |
| Named-human impersonation | Does the AI claim to be a specific named employee? |
| Call recording disclosure | Do you disclose recording where state law requires it (e.g., California, Illinois)? |
| Vendor contract | Does your AI vendor indemnify against disclosure failures? |
Call recording disclosure is a separate but related topic. Several states — including California, Illinois, Florida, and Michigan — require all-party consent to record a call. If your AI receptionist records conversations (most do, for quality review), you need a disclosure for that as well. Usually this is folded into the same opening statement: “This call may be recorded.”
The Business Case for Disclosing Voluntarily
If the law is unsettled, why disclose proactively rather than wait for clearer rules?
Three reasons.
Regulatory direction is clear even if the specific rules are not. Legislators and regulators across the country are moving toward mandatory AI disclosure in commercial contexts. Getting ahead of this now costs you one sentence. Getting caught later costs enforcement action and reputational damage.
Disclosure is also correlated with better call outcomes. Callers who know they’re speaking to an AI and proceed anyway have self-selected for engagement. They’re not waiting to be disappointed. Compare that to a caller who spends the whole conversation suspicious about what’s on the other end of the line.
Finally, it differentiates you from the bad actors. The businesses that have made AI calling look bad — scam calls, voice-cloned fraud — uniformly do not disclose. Standing on the other side of that line is a competitive advantage with customers who’ve been burned.
For a fuller look at how disclosure and transparency shape caller trust, see the pillar guide: Will Customers Hate Talking to an AI Receptionist?
What to Do Now
If you’re evaluating or already running an AI receptionist, audit the opening script today. Confirm it:
- Identifies itself as AI in the first sentence
- Leads immediately with what it can do for the caller
- Offers a path to a human if preferred
- Includes any required call-recording notice
If you’re considering AI Receptionist vs. Human Receptionist: Which One Do Callers Actually Prefer? — the short answer is that callers prefer whichever one solves their problem fastest. An AI that’s transparent and efficient wins that comparison most of the time.
The law on AI voice agents will continue to tighten. The businesses that build disclosure into their call flow now are building something the next wave of regulation will only confirm: that honesty is the right operating model, and it happens to also work.