An AI voice agent that answers or places calls for your business falls under the same calling laws as any automated call, plus a growing set of state rules on telling people they are talking to AI. The risk sits in three places: consent for outbound calls, disclosure that the caller is not human, and consent to record.
Short answer: for outbound calls, the FCC treats an AI-generated voice as an "artificial" voice under the Telephone Consumer Protection Act, so you need prior express consent, and prior express written consent for telemarketing. Every such message must identify your business and give a callback number. For disclosure, there is no single federal rule yet, but states such as Utah and Maine require it in some situations. The safest default is to say it is AI at the start of every call.
Outbound calls: the FCC treats AI voices as artificial
On February 8, 2024, the FCC announced a Declaratory Ruling that "recognizes calls made with AI-generated voices are "artificial" under the Telephone Consumer Protection Act (TCPA)." The ruling took effect immediately.
That puts AI voice calls under the TCPA rules for artificial or prerecorded voices. The statute bars calls to cell phones using "an artificial or prerecorded voice" without "the prior express consent of the called party," with limited exceptions such as emergencies. Similar limits apply to residential lines, subject to exemptions the FCC has set.
Telemarketing carries a higher bar. The FCC release notes that its rules require telemarketers "to obtain prior express written consent from consumers before robocalling them," and that the ruling "ensures AI-generated voices in calls are also held to those same standards."
What every AI voice message must say
FCC rule 47 CFR 64.1200(b) sets identification duties for all artificial or prerecorded voice messages. At the beginning of the message, the call must "state clearly the identity of the business, individual, or other entity that is responsible for initiating the call." During or after the message, it must state a telephone number for that business, other than the number of the system that placed the call.
Telemarketing messages to residential lines and cell phones must also offer an automated voice or key press opt out, within two seconds of identifying the caller, that adds the number to your do not call list. If your voice agent makes sales calls, build that opt out into the script.
Inbound calls and the question "am I talking to a robot?"
The TCPA rules above cover calls you place. When a customer calls you and an AI agent answers, the questions shift to disclosure and recording, and state law does most of the work.
| Rule | What it requires | Applies to voice? |
|---|---|---|
| FCC TCPA ruling (2024) | Consent for AI voice calls; written consent for telemarketing | Yes, outbound calls |
| 47 CFR 64.1200(b) | Name the business at the start; give a callback number | Yes, artificial voice messages |
| Utah Code 13-77-103 | Disclose AI if asked; proactive disclosure for high-risk regulated services | Yes, audio is covered |
| Maine 10 M.R.S. 1500-DD | Clear notice that the consumer is not dealing with a human, where they could be misled | Yes, covers aural communication |
| California B.O.T. Act | Disclose a bot used online to influence a purchase or vote | Written for online accounts; phone coverage unclear |
| California Penal Code 632 | Consent of all parties to record a confidential communication | Yes, telephone calls |
Utah. A supplier that uses generative AI in a consumer transaction must disclose that the person is "interacting with generative artificial intelligence and not a human, if the individual asks or otherwise prompts the supplier about whether artificial intelligence is being used." The question must be "a clear and unambiguous request." For regulated occupations, the duty is stronger: in a high-risk interaction, such as collecting health or financial data or giving financial, legal or medical advice, the disclosure must be made "verbally at the start of a verbal interaction." Utah's definition of generative AI includes systems that simulate conversation through audio.
Maine. A 2025 law bars using an AI chatbot or other technology in trade with a consumer "in a manner that may mislead or deceive a reasonable consumer into believing that the consumer is engaging with a human being unless the consumer is notified in a clear and conspicuous manner that the consumer is not engaging with a human being." Maine defines a chatbot to include "textual or aural communications," so it reaches voice. A violation counts under the Maine Unfair Trade Practices Act.
California. The B.O.T. Act, in force since July 1, 2019, makes it unlawful to use a bot to communicate with someone in California "online" with intent to mislead them about its artificial identity to drive a sale or influence a vote. Disclosure is a defense. The law defines "online" as a website or app and does not mention phone calls, so its reach to voice agents is unsettled.
Recording and transcription need consent too
Many voice agents record or transcribe calls so they can summarize them into a CRM. That triggers call recording laws. California Penal Code 632 punishes recording a confidential communication, including by telephone, "without the consent of all parties." A short notice at the start of the call, such as "this call is recorded and handled by an AI assistant," tells callers about both in one sentence. Ask counsel whether notice counts as consent in the states you call.
Federal AI disclosure: proposed, not final
In August 2024 the FCC proposed rules that would define an "AI-generated call" and require callers to disclose when a call uses an AI-generated voice, including at the point consent is collected. The proposal was published in the Federal Register on September 10, 2024. As of October 2026 we could not find a final order adopting it, and the current text of the FCC's calling rules does not mention artificial intelligence.
A related change has been undone. The FCC's one-to-one consent rule, aimed at comparison shopping and lead generation sites, was vacated by the Eleventh Circuit on January 24, 2025, in Insurance Marketing Coalition v. FCC, and sent back to the FCC.
A practical checklist
- Disclose at the start of every call that the caller is an AI assistant, and name your business. It meets Utah and Maine and costs one sentence.
- Answer honestly whenever a caller asks if they are talking to a person, and offer a transfer to one.
- For outbound calls, keep consent records per number, and use written consent for anything that markets a product.
- State a callback number in every outbound message and give telemarketing calls an in-call opt out.
- Add a recording notice wherever the agent records or transcribes.
- Check rules for your industry and states with counsel. This article is a plain-terms summary, not legal advice.
How Lean Discovery Group helps
Lean Discovery Group builds AI voice agents that answer calls around the clock, qualify callers, book appointments and transfer to a person when it matters, with every call summarized in your CRM. We script the real conversations, define clear handoff rules, and review call recordings, transcripts and outcomes after launch. Outbound reminders and follow ups are built within the calling rules for your industry.
Voice agents share knowledge with our AI chatbots and work alongside AI lead response automation. Book a discovery call to talk through your call flows.
Frequently asked questions
Do AI voice agents have to say they are AI?
There is no single federal rule yet. Utah requires disclosure when a consumer asks, and up front in some regulated services. Maine requires clear notice where a consumer could be misled into thinking they are talking to a human. Disclosing at the start of every call is the safest default.
Is it legal to make outbound calls with an AI voice?
Yes, with consent. The FCC ruled in 2024 that AI-generated voices are artificial voices under the TCPA, so calls need prior express consent, and telemarketing calls need prior express written consent.
What must an AI voice message say?
Under 47 CFR 64.1200(b), it must name the business responsible at the start and give a telephone number for that business during or after the message. Telemarketing calls to residential and cell phone lines also need an automated opt out.
Do I need consent to record AI phone calls?
Often, yes. California requires the consent of all parties to record a confidential call, and some other states have similar rules. A short notice at the start of the call is the usual approach, but confirm with counsel that it meets each state's rule.
Does the California bot disclosure law cover phone calls?
It is unclear. The B.O.T. Act applies to bots used online, defined as websites and apps, and does not mention phone calls. Many businesses disclose on calls anyway.
Sources
- FCC, "FCC Makes AI-Generated Voices in Robocalls Illegal," news release, February 8, 2024
- FCC, Declaratory Ruling FCC 24-17, CG Docket No. 23-362, released February 8, 2024
- 47 U.S.C. 227, Telephone Consumer Protection Act, U.S. Code via govinfo
- 47 CFR 64.1200, eCFR
- Federal Register, "Implications of Artificial Intelligence Technologies on Protecting Consumers From Unwanted Robocalls and Robotexts," proposed rule, September 10, 2024
- U.S. Court of Appeals for the Eleventh Circuit, Insurance Marketing Coalition v. FCC, No. 24-10277, January 24, 2025
- Utah Code 13-77-103, Disclosure requirements
- Utah Code 13-77-101, Definitions
- Maine Revised Statutes, Title 10, Chapter 241, including section 1500-DD
- California Business and Professions Code 17940 to 17943 (B.O.T. Act)
- California Penal Code 632
