You added an AI receptionist, and it is doing the job you gave it. It answers incoming calls, gathers the reason for the call, and helps keep routine requests moving. The next idea is natural: let the assistant call missed callers back, confirm an appointment, or follow up with a customer when a person is busy.
That next instruction changes the direction of the call. An assistant that answers is receiving a call. An assistant that dials is initiating one. For the FCC ruling discussed here, that is the practical line to mark before you build the callback.
Prefer to watch? See the short video version of this guide.
The FCC ruling in plain English
The FCC's Declaratory Ruling FCC 24-17, CG Docket No. 23-362, was adopted on February 2, 2024 and released on February 8, 2024. In that ruling, the FCC confirmed that the TCPA's restrictions on the use of an artificial or prerecorded voice encompass current AI technologies that generate human voices.
The same FCC ruling says calls using those AI technologies fall under the TCPA and the FCC's implementing rules. It says they require the prior express consent of the called party to initiate the calls, absent an emergency purpose or exemption. The ruling also states the point directly: callers must obtain prior express consent before making a call that uses an artificial or prerecorded voice simulated or generated through AI technology.
In its separate news release dated February 8, 2024, the FCC said the ruling takes effect immediately. That timing statement comes from the news release. The consent standard comes from the ruling itself.
Separate incoming from outgoing
A customer calling your number while an AI answers is a different question. FCC 24-17 does not address that situation, so this article draws no legal conclusion about it.
On the outbound side, the FCC ruling says the requirements apply to any AI technology that initiates any outbound telephone call using an artificial or prerecorded voice to consumers. A callback flow belongs on this side of your process map when the AI system starts the call.
So label direction before discussing consent. Write incoming beside a flow that begins when the customer calls you. Write outgoing beside a flow that begins when your AI assistant dials. Then review each outgoing flow under the FCC ruling instead of treating the whole phone system as one undivided feature.
Prior express consent comes before the call
For an outbound call covered by the ruling, the FCC uses the term prior express consent. The sequence matters in the agency's wording: the caller must obtain that consent from the called party before making the call that uses an artificial or prerecorded voice simulated or generated through AI technology.
The FCC ruling sets a different standard when the AI voice call introduces an advertisement or contains telemarketing. In that situation, the FCC's rules require the caller to obtain the called party's prior express written consent. Keep that full phrase in your workflow notes rather than reducing it to informal shorthand.
For a small-business review, this creates a useful separation. First identify whether the AI will initiate an outbound call. Then identify whether that call introduces an advertisement or contains telemarketing. Record the applicable FCC term beside the flow, and have a lawyer check how the term applies to your actual facts before the dialing step goes live.
Identify the responsible caller
Consent is not the only item in the FCC ruling. The ruling says the FCC's rules require all artificial or prerecorded voice messages to provide certain identification and disclosure information for the entity responsible for initiating the call. It also states that, at the beginning of the message, the identity of the responsible business, individual, or other entity must be stated clearly.
The FCC ruling adds another requirement when the artificial or prerecorded voice message includes or introduces an advertisement or constitutes telemarketing. The message must offer the specified opt-out methods so the called party can request that calls to that telephone number stop.
Put those items directly into the call script review rather than leaving them in a separate note. Start with the identification line. If the call includes or introduces an advertisement or constitutes telemarketing, place the required opt-out method in the same reviewed flow. This is a practical way to keep the script aligned with the two requirements the FCC describes.
A person behind the AI is not a carve out
Owners sometimes design the assistant so a person can take over, approve an exception, or monitor the conversation. That may be useful for the business, but the ruling points the other way on a related argument. It says the TCPA does not allow for any carve out of technologies that purport to provide the equivalent of a live agent, and it relies on its earlier Soundboard ruling: a live agent who selects the prerecorded messages to be played does not negate the prohibition on initiating a call with an artificial or prerecorded voice.
A human handoff can still be part of a careful workflow. It should not be used as the reason to classify the AI-generated voice as outside this ruling. For the consent check, focus on what initiates the outbound call and whether the call uses an artificial or prerecorded voice, the features the FCC names.
Interested in implementing similar AI solutions? Discover how PATech Labs can help your business leverage cutting-edge artificial intelligence.
Learn About Our ServicesWhat the FCC news release adds
On February 8, 2024, the FCC's news release announced the unanimous adoption of a Declaratory Ruling recognizing calls made with AI-generated voices as artificial under the Telephone Consumer Protection Act. That is the agency's description of what it had adopted.
The same FCC news release says the TCPA gives the FCC civil enforcement authority to fine robocallers. That statement identifies who can act under the statute. It does not tell any individual business whether a particular call is legal or illegal.
The release headline says, FCC Makes AI-Generated Voices in Robocalls Illegal. Keep that headline attributed to the release. Do not turn it into a claim that AI calls are banned. The ruling text gives the more precise rule used in this article: covered calls require prior express consent absent an emergency purpose or exemption, with prior express written consent required when the calls introduce an advertisement or contain telemarketing.
The later disclosure rule was proposed
On September 10, 2024, the FCC published a proposal that, according to the Federal Register notice, would define AI-generated calls and adopt rules requiring callers to disclose to consumers when they receive an AI-generated call. The Federal Register presents this as a proposed rule, not an adopted rule.
As of October 8, 2026 we found only a proposed rule, not a final one; check the FCC site for updates. Our check used the Federal Register because www.fcc.gov could not be read. This is a statement about what that search found, not a claim that the proposal can never change.
Do not mix that proposal with the requirements already described in FCC 24-17. The proposal concerns a disclosure for an AI-generated call. The adopted ruling separately says all artificial or prerecorded voice messages must provide certain identification and disclosure information for the entity responsible for initiating the call. One is a proposal, while the other appears in the ruling.
A practical review for this week
Start with a simple inventory of every voice flow. Mark each one incoming or outgoing based on who starts the call. Keep the inbound flows visible, but note that FCC 24-17 does not address a customer calling in while an AI answers. Move every flow in which the AI dials a consumer into the outbound review.
For each outgoing flow, write down where the prior express consent comes from. If the call introduces an advertisement or contains telemarketing, write down where the prior express written consent comes from. Keep the record with the flow description so the consent check is connected to the exact dialing action you are reviewing.
Next, review the opening words for clear identification of the entity responsible for initiating the call. For a message that includes or introduces an advertisement or constitutes telemarketing, verify the specified opt-out method. These steps mirror the identification and opt-out requirements stated in the FCC ruling without trying to decide a specific case in the abstract.
Build consent first, then call
A straightforward workflow can stop before dialing and check a consent field tied to the exact call flow your lawyer reviewed. If the expected record is present, the process can continue to the reviewed call script. If it is missing, unclear, or does not match the kind of call being made, the process can pause instead of letting the assistant make the decision.
Send that exception to a person for approval. The person can check the record, the purpose of the call, and the script before deciding what happens next. This human review is an operational control. Under the ruling's Soundboard reasoning, a live agent in the loop does not take the call outside the rule.
PATech Labs builds AI automation for US small businesses, including reminders and verification steps that route exceptions to a person for approval. That is one way to place a consent check before an AI callback dials. A written checklist used consistently does the same job.
Keep the review grounded
This article provides information, not legal advice, and it does not decide whether your specific calling setup is legal or illegal. Ask a lawyer to review your own consent records, call purposes, scripts, and exceptions against the rules that apply to your situation.
For adjacent context, read our guide to written-consent rules for texting and calling customers and our AI receptionist compliance overview for US small businesses. Neither article discusses FCC 24-17 or the ruling explained here, so treat them as related reading rather than sources for this analysis.
The line to remember
An AI that answers a customer's incoming call and an AI that dials a customer are two different directions. FCC 24-17 addresses the second direction when AI technology initiates an outbound telephone call using an artificial or prerecorded voice to consumers. It does not address the first situation.
That is the line where the consent review starts. Name the direction of every flow, locate the required consent before an outbound AI voice call, include the identification and any applicable opt-out method, and route uncertainty to a person. The difference between an AI that answers and an AI that dials is small on a workflow diagram, but it is the central distinction in this ruling.