Is AI Calling Legal for Insurance Agencies?

4:12 PM on a Tuesday. A purchased auto lead lands in your pipeline, and your AI agent dials it ninety seconds later. The call goes well. The prospect books a quote appointment.

Then your office manager asks the question nobody on the team can answer: did we have consent to make that call with an AI voice, and can we prove it?

That question has a real answer, and it changed recently. Here is what the rules actually say for a US insurance agency running AI calls and texts, what is new on November 25, 2026, and what a compliant setup looks like in practice.

This is general information for insurance agency owners, not legal advice. TCPA exposure is specific to your agency, your lead sources and your states. Have your own counsel review your consent language and calling practices before you switch anything on.

Does the TCPA apply to an AI voice calling my insurance leads?

Yes. According to the Federal Communications Commission's Declaratory Ruling FCC 24-17, adopted February 2, 2024, the TCPA's restrictions on an "artificial or prerecorded voice" cover current AI technologies that generate human voices. The ruling is explicit that there is no carve out for technologies that behave like a live agent.

That matters more than it first appears. An AI agent that holds a genuine back and forth conversation, handles objections and sounds entirely natural is still an artificial voice in the eyes of the law. The sophistication of the agent is not a defense. The ruling covers both fully synthetic voices and voices cloned from a real person.

The practical consequence for an agency: every outbound AI call carries the same obligations as a prerecorded robocall, including consent, caller identification and opt-out handling.

What consent does an outbound AI call to an insurance lead need?

It depends on what the call is for, and the distinction is the one most agencies get wrong. The requirements sit in 47 C.F.R. section 64.1200.

A quote solicitation is a sales call. If your AI agent is calling a purchased lead to sell a policy, you are in prior express written consent territory, not the lighter informational standard.

Written consent is not a checkbox buried in a footer. The agreement needs to identify the phone number being authorized, state clearly that the consumer is authorizing artificial or prerecorded voice calls, and say that agreeing is not a condition of purchase.

One caution on purchased leads. Consent travels with the specific disclosure the consumer actually saw at the point of capture. A vendor telling you a list is "TCPA compliant" is not the same as you holding consent language that names artificial voice calls. Ask to see the actual form.

What changes on November 25, 2026?

This is the part most agencies have not caught yet. On September 30, 2026 the FCC adopted a Report and Order, FCC 26-67, revising how consumers revoke consent. It was published in the Federal Register on October 26 at 91 FR 67214 and takes effect November 25, 2026, codified at 47 C.F.R. section 64.1200(a)(10).

Three changes matter for an agency running AI campaigns.

01
You may now designate an exclusive opt-out method.

Under the previous rules a consumer could revoke consent by any reasonable means and you were barred from restricting how. Starting November 25, a caller may designate one or more FCC-approved methods as the exclusive way to revoke, provided that method is clearly and conspicuously disclosed in every robocall that requires consent. If you do not designate one, you must keep honoring revocation requests made by any reasonable means.

02
Informational opt-outs can be read narrowly.

A revocation directed at one category of informational call can now be treated as applying to that category only, rather than stopping every communication that requires consent.

03
Telemarketing opt-outs still stop everything.

For sales calls and marketing texts, a single revocation request opts the consumer out of all future telemarketing from or on behalf of your agency. The narrowing applies to informational communications, not to selling.

The runway is short. The FCC rejected a request for a twelve month implementation period. Revocations still have to be honored within a reasonable time not exceeding ten business days, and your systems need to recognize the standardized opt-out words.

Do inbound AI receptionists need consent too?

No, and this is the cleanest distinction in the whole framework.

The TCPA governs calls you initiate. The FCC has stated that its voice requirements do not extend to technologies used to answer inbound calls. When a customer dials your agency and an AI answers, the consent analysis that applies to outbound calling simply does not arise.

For an agency weighing where to start with AI, that asymmetry is worth planning around. Answering every inbound call carries a materially lighter compliance load than working outbound lists, and it is the faster thing to switch on.

What about SMS and A2P 10DLC?

Texting adds a second, separate requirement that has nothing to do with the FCC.

A2P 10DLC is carrier registration. Before any business can send application-to-person messages over standard ten digit numbers, US carriers require a registered and approved campaign. Approval depends on your privacy policy, your terms of service and your lead forms capturing explicit consent to receive marketing messages, plus sample messages that match what you actually send.

Two things agencies underestimate. Approval takes real elapsed time, historically weeks rather than days. And the consent language carriers want to see is often missing from exactly the same lead forms that fail the TCPA written consent test above. Fixing the forms once solves both problems.

Email has its own rules. Under CAN-SPAM every commercial message needs a valid physical postal address and a working unsubscribe mechanism, and opt-outs must be honored.

What does a compliant AI calling setup actually look like?

Pulling it together, here is the checklist an insurance agency needs before an AI agent dials anyone.

Where this fits

Agencies with clean, recent, first party leads and a consent form written after 2024. Agencies starting with inbound, where the consent question does not arise. Agencies with the operational discipline to maintain suppression lists properly.

Where this fails

Aged purchased lists with no visibility into the original consent language. Agencies that treat a vendor's compliance assurance as their own defense. Anyone planning to send SMS next week, because A2P 10DLC approval will not be ready. And any setup where opt-outs are handled manually, which breaks the ten business day requirement the moment volume rises.

How SUPERAGENT handles this

Disclosure: SUPERAGENT is an AI Business Partner built only for US insurance agencies, so we are one of the options in this category.

The compliance layer is a large part of what the platform is. SUPERAGENT operates as a Twilio reseller and creates a direct customer sub account for each agency, running SHAKEN/STIR attestation and voice integrity work so carriers treat the traffic as legitimate. Numbers are checked against Do Not Call registries before dialing. Calling windows run 8 AM to 9 PM with state by state variation already handled, Sunday calling and texting is hard blocked, and Saturday is treated as a gray zone the agency can override.

A2P 10DLC registration is submitted and iterated on the agency's behalf, though the agency still has to update its own privacy policy, terms of service and lead forms to capture explicit consent. Every outbound email automatically inserts the agency's physical address and an unsubscribe link, and clicking it removes the lead from the campaign. Opt-outs are honored automatically, and a lead who opts out is visibly blocked from further outreach. SUPERAGENT discloses that the call uses AI at the start of every voice call.

Frequently asked questions

Is AI calling legal for insurance agencies?

Yes, with consent. AI voice calls are not banned. Under FCC 24-17 they are treated as artificial voice calls under the TCPA, which means outbound calls need prior express consent, and sales calls need prior express written consent.

Do I need written consent to call a purchased insurance lead with AI?

If the call is a quote solicitation, yes. Telemarketing calls using an artificial voice to a mobile or residential line require prior express written consent that names the number and states that consent is not a condition of purchase.

Does an AI receptionist answering calls need TCPA consent?

No. The FCC has said its voice requirements do not extend to technologies used to answer inbound calls. The consent framework applies to calls you initiate.

What is changing in the TCPA on November 25, 2026?

FCC 26-67 takes effect. Callers may designate an exclusive method for revoking consent if it is clearly disclosed in every robocall requiring consent, and revocations of informational calls may be read as category specific. Telemarketing revocations still stop all future telemarketing.

How long do I have to honor an opt-out?

A reasonable time not exceeding ten business days.

Do I have to tell people they are talking to an AI?

The FCC has proposed requiring a clear disclosure at the beginning of each call using an AI-generated voice, but that proposal has not been adopted. Identification and disclosure obligations for artificial voice calls apply today, and some states impose their own disclosure rules. Disclosing at the start of the call is the safer practice.