The Short Answer: Yes, But Only With Consent

AI marketing calls are legal under the TCPA when prior express consent is obtained. The FCC’s February 2024 Declaratory Ruling (FCC-24-17) explicitly confirmed that AI-generated voices are treated as "artificial or prerecorded voice" under the TCPA, subjecting them to the same consent requirements as traditional robocalls. This means conversational capability does not remove robocall rules, regardless of how human the voice sounds.

Non-compliance carries significant financial risk, with TCPA statutory damages ranging from $500 to $1,500 per call and no aggregate cap. Recent class-action settlements have exceeded $925 million in aggregate, underscoring the financial exposure for businesses that fail to secure proper consent. These penalties apply per violation, making even small campaigns potentially costly if consent records are incomplete or outdated.

For marketing calls to U.S. cell phones, prior express written consent (PEWC) is required in 47 states, while informational or transactional calls may only require prior express consent (PEC), which can be oral. However, regulatory volatility persists: the Eleventh Circuit vacated the FCC’s one-to-one consent rule in January 2025, and state-level laws now create a patchwork of requirements. In Texas, Louisiana, and Mississippi, oral consent may suffice for marketing AI calls following the Bradford v. Sovereign Pest Control ruling, but this exception is narrow and does not authorize abandoning written consent for nationwide programs.

• Obtain prior express written consent (PEWC) for all marketing AI calls to U.S. cell phones, except in Texas, Louisiana, and Mississippi where oral consent may apply post-Bradford.
• Engineer compliance into infrastructure with real-time consent linkage, automated DNC suppression (updated at least every 31 days), and opt-out honoring within 10 business days.
• Provide clear, immediate AI disclosure at the start of every call, adhering to state-specific timing (e.g., within 30 seconds in Texas) and allowing consumers to request a human or opt out via keywords.

At My AI Call Center, we build compliance into every campaign by reviewing list source and consent records before launch, ensuring only approved, permissioned, or reviewed lists are used. Our process includes verifying consent disclosure of AI use and honoring opt-outs immediately—never inventing metrics or skipping validation steps. This disciplined approach helps multi-location organizations run useful calls that confirm, qualify, remind, survey, retain, and connect without exposing themselves to avoidable TCPA liability.

Why the Rules Keep Moving: Federal Rulings, Court Vacaturs, and State Laws

Just when a compliance team maps the TCPA rules, the rules move. That is not bad luck — it is the defining feature of this legal landscape, and it catches businesses off guard every year.

The clearest example is the FCC's one-to-one consent rule. Adopted in December 2023, it would have required consent to be tied to a single identified seller, with calls "logically and topically associated" with the interaction that prompted the consent. It never took effect: the Eleventh Circuit vacated the rule in January 2025, leaving companies that had rebuilt their consent flows scrambling to interpret what still applies. As one industry analysis put it, the lesson is not that consent requirements relaxed — it is that the rules keep moving, expanded one year and vacated the next.

State courts add another layer. The Fifth Circuit's ruling in Bradford v. Sovereign Pest Control created a narrow exception permitting oral consent for marketing AI calls in Texas, Louisiana, and Mississippi. But experts caution that the defensive value of Bradford is real yet narrow: it helps operators who capture verbal consent during inbound calls and log it cleanly. It does not authorize anyone to abandon written consent capture for nationwide outbound programs.

State legislatures are filling the gaps the federal government leaves open:

• Texas requires AI disclosure within 30 seconds of a call, with California, Florida, Colorado, Illinois, and Utah enforcing their own variants.
• Florida, Oklahoma, and Texas have telemarketing laws stricter than the federal TCPA.
• Roughly half of U.S. states maintain their own calling-time and holiday rules on top of federal requirements.

A federally mandatory AI disclosure is also expected through a pending FCC rulemaking, so today's state patchwork may soon sit atop a national floor.

Two assumptions trip up even careful organizations. First, an established business relationship does not exempt AI calls from TCPA consent requirements — it only relieves callers from DNC list restrictions. Second, B2B calls are not categorically exempt: while many calls to businesses escape the FTC's Do Not Call provisions, that exemption does not resolve TCPA wireless restrictions, state mini-TCPA laws, or recording requirements. A personal cell number is a personal cell number, even when the owner runs a business.

This volatility is why static checklists fail. Compliance has to be engineered into infrastructure — consent linkage, disclosure scripts, and opt-out handling that adapt as rulings change. My AI Call Center builds campaigns against approved, permissioned, or reviewed lists only, checking list source and consent records before launch, because in an environment where the rules shift mid-year, the safest position is a documented, auditable one.

The Consent Standard: What Counts as Legal Permission for an AI Call

Consent is the load-bearing wall of every legal AI calling program — and under the TCPA, "the person seemed friendly on the phone" does not count as a foundation. Get this piece wrong, and statutory damages of $500 to $1,500 per call stack up with no aggregate cap, which is why class actions, not the FCC, pose the bigger threat to most brands.

For marketing calls to U.S. cell phones, the standard is prior express written consent (PEWC) in 47 states. E-signatures satisfy the writing requirement under the E-SIGN Act, and any promotional content — even a light upsell inside an otherwise transactional call — classifies the call as marketing, raising the bar to written consent.

The Bradford v. Sovereign Pest Control ruling changed the map slightly. Texas, Louisiana, and Mississippi now permit oral consent for marketing AI calls, but legal analysts describe Bradford's defensive value as "real but narrow" — it helps operators who capture verbal consent during inbound calls and log it cleanly. It does not authorize abandoning written consent capture for nationwide outbound programs.

Consent language itself must name the technology. Compliance guidance is explicit that consent requires clear disclosure of AI use, with sample language referencing "an AI-generated voice." A consent form that mentions only "calls" or "communications" leaves a gap.

Consent also does not travel across channels. Consent obtained for voice calls does not automatically extend to text messages unless the original disclosure explicitly covers both. Multi-touch campaigns that follow a call with a text need their own consent basis for each channel.

A workable consent record holds four elements:

• Clear disclosure that calls may use an AI-generated voice
• The identified seller on whose behalf calls are made
• A timestamp and link to the specific dialed number
• Retention for at least five years — beyond the TCPA's four-year statute of limitations, as defense counsel recommend for record retention

Then there is the liability reality. The entity on whose behalf the calls are made bears liability, regardless of which downstream vendor pressed dial — a point underscored by Lamb v. Mortgage One Funding, which defined the class to include calls made by vendors, lead generators, and agents alike. Outsourcing the dialing does not outsource the exposure.

That is why list discipline matters as much as script quality. My AI Call Center reviews list source and consent records before any campaign launches and flags bought lists without clear permission records — because under the TCPA, the brand name on the call owns the outcome.

Engineering Compliance Into the Campaign, Not a Checklist

Engineering Compliance Into the Campaign, Not a Checklist

Effective TCPA compliance for AI outbound calls cannot be bolted on as an afterthought—it must be woven into the technical and operational fabric of every campaign. This means moving beyond static policies to real-time systems that enforce consent, honor opt-outs, and maintain audit trails without relying on manual intervention. The stakes are high: TCPA statutory damages range from $500 to $1,500 per call, and aggregate class-action verdicts have exceeded $925 million across recent dockets.

To survive legal scrutiny, infrastructure must enforce specific, measurable safeguards. Consent must be linked in real time to the dialed number so the system can verify authorization before initiating contact. Do Not Call (DNC) lists require scrubbing at least every 31 days against the National DNC Registry, as numbers are added daily and batch processing creates unacceptable risk. Opt-out requests—whether via keywords like "STOP" or verbal revocation—must be honored within 10 business days, and call abandonment rates must stay under 3% of answered calls to avoid triggering additional liability.

• Real-time consent linkage ensures prior express consent is validated at call time, not assumed from a static file.
• DNC scrubbing every 31 days aligns with federal safe harbor requirements and state-level expectations.
• Opt-out honoring within 10 business days meets both TCPA and FTC Telemarketing Sales Rule standards.
• Abandonment under 3% of answered calls avoids triggering predictive dialer restrictions under TCPA.
• AI disclosure and human-escalation options on every call satisfy emerging state mandates and prepare for federal NPRM requirements.
• Retaining consent records and call logs for at least five years exceeds the four-year TCPA statute of limitations, preserving evidence for class-action discovery.

For organizations using managed services like My AI Call Center, this infrastructure approach transforms compliance from a liability into a operational advantage. By embedding these controls into campaign design—from script approval to outcome routing—businesses can run structured AI calling campaigns that confirm, qualify, or retain contacts without exposing themselves to avoidable risk. The focus shifts from hoping compliance is followed to knowing it is enforced, call by call.

How to Launch a Compliant AI Calling Campaign

Knowing the rules is one thing. Building a launch process that follows them — every campaign, every call — is where most organizations actually get into trouble.

Start with one clear goal. Decide what the call needs to accomplish — confirm an appointment, qualify a lead, renew a member — and scope the campaign around that single outcome. A focused campaign is easier to script, easier to disclose, and far easier to defend if a regulator or plaintiff's counsel comes asking questions later.

Next, verify your list before you spend a dollar. Review the list source, confirm consent records exist, and check whether consent covers AI voice calls specifically — consent language must clearly disclose the use of an AI-generated voice. If a bought list arrives without clear permission records, decline it. The stakes are simply too high: TCPA statutory damages run $500 to $1,500 per call with no aggregate cap, and 2025–2026 class-action settlements have ranged from $4.75 million to $9.95 million.

Then approve the script and disclosure language before anything launches. Your disclosure should state the call is AI-assisted, identify who is calling, and offer a path to a human or an opt-out — a sample disclosure pattern opens with the AI notice and purpose in the first breath. State timing rules matter too: Texas requires disclosure within 30 seconds, with California, Florida, Colorado, Illinois, and Utah enforcing their own variants.

When the campaign goes live, run calls inside approved windows and monitor outcomes in real time. The federal window is 8:00 AM to 9:00 PM in the recipient's local time, but roughly half of states impose their own calling-time rules. Real-time monitoring matters because numbers hit the National DNC Registry daily, and DNC scrubbing is required at least every 31 days — a weekly suppression sync is not fast enough.

Finally, route outcomes back with a full audit trail:

• A dispositioned contact list with codes — confirmed, qualified, renewed, opted out, no answer
• Opt-out and DNC logs, honored immediately and carried into your permanent DNC records
• Per-call notes and follow-up requests routed into your CRM and scheduling tools
• A completion and coverage report showing what actually happened

Keep those records well beyond the campaign: the TCPA's statute of limitations runs four years, and best practice is retaining consent records and call logs for at least five years. Remember that liability follows the entity on whose behalf calls are made, regardless of which vendor dialed the phone. This is exactly why My AI Call Center checks list source and consent records before any campaign launches — and tells you plainly if the list won't support the campaign. One note: requirements vary by location, industry, and consent status, so obtain qualified legal guidance before launch.