Why the Autodialer Question Still Confuses Callers

Every call placed without a compliant system is a potential lawsuit. TCPA statutory damages run $500 to $1,500 per call, and with a four-year statute of limitations, a single campaign can add up fast. That math is why so many calling teams assume any bulk dialing setup counts as an autodialer. The assumption is understandable, and it's often wrong.

The statutory definition lives at 47 U.S.C. § 227(a)(1). It covers equipment with the capacity to store or produce telephone numbers using a random or sequential number generator, and to dial those numbers. That phrase, "using a random or sequential number generator," is where nearly all the confusion comes from.

For years, courts couldn't agree on what it modified. The Third, Seventh, and Eleventh Circuits read it narrowly, applying the generator requirement to both "store" and "produce." The Second, Sixth, and Ninth Circuits read it broadly, applying it only to "produce." Under the broad reading, almost any system that dials from a stored list qualified as an autodialer, which threatened to sweep in ordinary smartphones and everyday dialing equipment. The D.C. Circuit had already warned in its 2018 ACA International decision that the FCC's expansive "potential functionalities" reading created an "eye-popping sweep."

The Supreme Court resolved the split on April 1, 2021, in Facebook, Inc. v. Duguid. Unanimously, the Court held that the random or sequential number generator requirement modifies both "store" and "produce." The shorthand version: no random or sequential number generator, no autodialer. Equipment that merely stores and dials numbers from a curated list falls outside the definition, no matter how many calls it places.

Two practical points follow from that ruling:

• Volume alone doesn't decide anything. The FCC's June 25, 2020 Declaratory Ruling made clear that calling a large volume of numbers is "not probative" of autodialer status.
• Predictive dialers generally fall outside the definition unless they have the present capacity to store and produce numbers randomly or sequentially.
• Consent still matters most. Even if your equipment qualifies as an autodialer, prior express consent from the recipient can shield you from liability entirely.

The narrower definition is good news, but it isn't a free pass. The TCPA's separate prohibition on artificial or prerecorded voice calls was left untouched by Duguid, and the FCC's February 8, 2024 ruling classified AI-generated voices as "artificial or prerecorded voice." That means an AI-assisted calling campaign needs proper consent even when the dialing equipment isn't an ATDS.

This is the lens we apply at My AI Call Center before any campaign launches. We review list source and consent records first, and we run structured campaigns against approved, permissioned, or reviewed lists only. If a bought list lacks clear permission records, we flag it, and in most cases we decline it. It's a simpler compliance posture than arguing about dialing equipment after the fact: know whose numbers you're calling, and be able to show why you're allowed to call them.

The Supreme Court's Answer: Facebook v. Duguid

For years, courts couldn't agree on what an autodialer was, and businesses paid the price. The Supreme Court finally settled the question on April 1, 2021, with a unanimous 9-0 decision in Facebook, Inc. v. Duguid, and the answer narrowed the TCPA's reach considerably.

The case turned on a comma. The statute at 47 U.S.C. § 227(a)(1) defines an autodialer as equipment that can "store or produce telephone numbers to be called, using a random or sequential number generator." Writing for the Court, Justice Sotomayor applied the series-qualifier canon and held that "using a random or sequential number generator" modifies both "store" and "produce." As Quarles & Brady summarized, the rule is simple: no random or sequential number generator, no autodialer.

That means equipment that merely stores a list of numbers and dials them does not qualify. The Court also rejected the broad reading adopted by the Ninth, Sixth, and Second Circuits, which had threatened to sweep in virtually every smartphone. Expanding the definition that far, the Court warned, "would take a chainsaw to these nuanced problems when Congress meant to use a scalpel."

The FCC had already moved in the same direction. In its June 25, 2020 Declaratory Ruling, the agency clarified that calling a large volume of numbers is "not probative" of autodialer status. The dispositive question is whether the equipment can dial random or sequential numbers without human intervention. Call volume alone doesn't tell you anything.

This matters for anyone running structured outbound campaigns. Equipment that dials from a curated, approved list, where a human reviewed the contacts and the numbers came from consent records rather than a generator, generally falls outside the ATDS definition. The practical takeaways:

• A device qualifies as an ATDS only if it uses a random or sequential number generator to store or produce numbers.
• High call volume, by itself, proves nothing about autodialer status under the FCC's 2020 ruling.
• Dialing from a fixed, human-reviewed list generally places equipment outside the definition.
• Prior express consent still shields you from TCPA liability even if an autodialer is involved.

One caution: Duguid left the TCPA's separate prohibition on artificial or prerecorded voice untouched, so consent obligations can apply independently of how the call was dialed. That's why list discipline matters as much as equipment choice. At My AI Call Center, campaigns run only against approved, permissioned, or reviewed lists, with source and consent records checked before launch, because the defensible position is a reviewed list plus documented consent, not a technicality about the dialer.

With statutory damages of $500 to $1,500 per call, the stakes are too high to guess. The generator question is now the legal test, and businesses that can document how their equipment actually dials hold a much stronger hand.

Where Duguid Stops: AI Voices and Separate Consent Rules

The Duguid decision narrowed what counts as an autodialer, but it didn't touch the TCPA's separate ban on artificial or prerecorded voices. That gap matters if your campaigns use AI-generated speech, because the consent obligation attaches to the voice itself, not to whether your dialing equipment meets the ATDS test.

The Supreme Court explicitly left the artificial or prerecorded voice prohibition untouched, meaning that rule applies no matter what technology places the call. The FCC's February 8, 2024 ruling then closed the obvious loophole by classifying AI-generated voices as "artificial or prerecorded voice" under § 227(b), reasoning that the legal status of a voice depends on how it is produced, not how human it sounds. An AI voice triggers prior express consent requirements on its own, even when the dialer is not an ATDS. The FCC was direct that the statute allows no carve-out for technologies that mimic a live agent.

Consent revocation rules changed again in 2026. Under the FCC's September 30, 2026 Report and Order, callers may designate exclusive opt-out methods, and the old "revoke all" rule gave way to category-specific revocations. Seven standardized words now count as opt-out requests:

• stop
• quit
• end
• revoke
• opt out
• cancel
• unsubscribe

The FCC's revised revocation framework gives callers up to ten business days to honor a request and requires clear, conspicuous disclosure of any designated opt-out method. Category-specific revocations mean a contact who opts out of one type of call hasn't necessarily opted out of every campaign you run, so your records need to track scope, not just status.

For multi-state programs, written consent remains the safer baseline. A Fifth Circuit ruling in February 2026 held that oral consent can satisfy the TCPA for automated or prerecorded telemarketing calls, but that decision binds only Texas, Louisiana, and Mississippi and doesn't displace state law. Statutory damages still run $500 to $1,500 per call, which is why My AI Call Center treats prior express consent as the anchor for every campaign rather than a checkbox.

That's the practical takeaway: confirm your consent records cover the voice you're using, log opt-outs across every campaign, and carry DNC requests into client records immediately. The ATDS question is only half the analysis.

How to Run Compliant Campaigns Regardless of the Label

The label on your dialer matters less than how you handle the data behind it. You can run compliant campaigns without obsessing over technical classifications, provided you anchor every call in verified consent.

Start by documenting your equipment’s technical characteristics. The Supreme Court’s ruling in Facebook, Inc. v. Duguid clarified that a device only qualifies as an autodialer if it uses a random or sequential number generator to store or produce numbers. Legal analysis confirms that equipment dialing from a curated, reviewed list generally falls outside this definition. If your system dials from a specific, permissioned list rather than generating random numbers, you have strong technical documentation to support your compliance posture.

Consent remains your most reliable shield. Even if your equipment technically meets the autodialer definition, obtaining prior express consent avoids TCPA liability. For multi-state campaigns, treat prior express written consent as the baseline. While the Fifth Circuit’s 2026 Bradford ruling relaxed federal requirements for oral consent in Texas, Louisiana, and Mississippi, state laws often still mandate written permission. Snell & Wilmer advises that written consent is the safest course for nationwide programs because it satisfies both federal regulations and varying state-law requirements.

You must also respect opt-outs rigorously. The FCC’s September 2026 Report and Order allows callers to designate exclusive opt-out methods, but you must honor those requests within ten business days. Standardized words like "stop" or "revoke" trigger immediate action. My AI Call Center reviews list sources and consent records before any campaign launches. We flag bought lists without clear permission records and decline them. This ensures we never spend your budget on a list that cannot support the campaign legally.

To maintain this discipline, keep consent records tied to every contact. This practice protects you against statutory damages, which range from $500 to $1,500 per call. Industry data shows that TCPA class-action filings have risen sharply, making record-keeping essential.

Practical steps for compliant execution include:

• Document that your dialer uses a set list, not a random generator.
• Capture prior express written consent for all multi-state outreach.
• Honor opt-outs within ten business days and update client DNC logs.
• Verify AI voice disclosure, as AI voices trigger separate consent rules.

The FCC’s 2024 ruling classified AI-generated voices as "artificial or prerecorded voice," triggering consent requirements independent of the dialer's status. Compliance guides note that the legal status depends on how the voice is produced, not how human it sounds. You must disclose AI assistance on every call and allow recipients to request a human.

Finally, ensure your campaign structure supports these obligations. A managed service should review the list source and consent records before launch. This pre-launch check identifies gaps in permission records immediately. You avoid the risk of calling a number that lacks valid consent. It keeps your operations clean and your legal exposure low.

Run more useful calls without building a bigger call center. Plan your campaign with a focus on approved, permissioned lists. We quote the full cost before launch, so you know exactly what you are paying for. No invented numbers, just clear outcomes.