The Autodialer Definition Problem That Cost Businesses Millions

Every outbound call carries a built-in price tag. Under the TCPA, statutory damages run from $500 to $1,500 per call, per violation, which is exactly why the statute has fueled so much high-volume class action litigation. When you multiply that per-call exposure across a campaign of thousands of numbers, the stakes of one legal definition become obvious.

The problem, until recently, was that nobody could tell you what that definition actually covered.

For years, courts disagreed. The Second, Sixth, and Ninth Circuits read the autodialer definition broadly, while the Third, Seventh, and Eleventh Circuits read it narrowly. If your call center sat in one federal circuit, your dialing equipment might qualify as an autodialer. The same equipment one state over might not. Compliance became a geography problem.

The FCC made it worse in 2015. Its Declaratory Ruling suggested that "capacity" included potential functionalities and even future possibilities, an interpretation so broad that the D.C. Circuit struck it down in ACA International v. FCC (2018), calling it an "eye-popping sweep" that could effectively cover every smartphone in America. Think about that: a ruling meant to clarify the law instead made ordinary cell phone use a potential TCPA violation.

Justice Sotomayor later captured the absurdity in Facebook, Inc. v. Duguid, writing that an expansive reading would "take a chainsaw to these nuanced problems when Congress meant to use a scalpel."

The uncertainty created real business paralysis:

• Companies couldn't predict whether standard dialing software exposed them to class action risk.
• Identical technology produced different legal outcomes depending on the circuit.
• Legitimate outreach, like appointment reminders and renewal calls, carried litigation risk comparable to actual spam.
• Plaintiffs' attorneys could plausibly argue almost any dialing system qualified.

This is the backdrop you need before the definition itself makes sense. Congress, as analysis of the Supreme Court's ruling points out, originally targeted a specific type of telemarketing equipment: gear that risks dialing emergency lines randomly or tying up all the sequentially numbered lines at a single business. It was never meant to regulate every automated calling tool.

That distinction matters for how you run outbound campaigns today. At My AI Call Center, for example, every campaign runs against approved, permissioned, or reviewed contact lists, with consent records checked before launch, because where your numbers come from shapes your legal exposure far more than how fast you dial them.

So what finally settled the question? In 2021, the Supreme Court answered it unanimously. The next section walks through that ruling and the definition it left behind.

The Duguid Standard: What Actually Makes a Device an Autodialer

For years, businesses feared that any automated dialing could trigger TCPA liability, with statutory damages of $500 to $1,500 per call fueling waves of class action litigation. That fear was largely put to rest in 2021, when the Supreme Court issued a unanimous 9-0 decision in Facebook, Inc. v. Duguid, authored by Justice Sotomayor.

The Court held that the phrase "using a random or sequential number generator" modifies both verbs in the statutory definition of an automatic telephone dialing system: "store" and "produce." That grammatical point carries real weight. A device only qualifies as an autodialer if it can store or produce numbers using a random or sequential generator.

The practical result: dialing a customer-supplied, pre-compiled list is not autodialer activity, even if the system dials those numbers automatically. Congress, the Court reasoned, targeted a specific type of equipment that risks dialing emergency lines randomly or tying up sequentially numbered lines at a single entity, not every modern calling tool. Justice Sotomayor warned that a broader reading would "take a chainsaw to these nuanced problems when Congress meant to use a scalpel," potentially sweeping in ordinary cell phones.

Lower courts have since fallen in line. In 2024, the Third Circuit's Perrong ruling reaffirmed the narrow standard, rejecting the argument that using a number generator to select from a previously compiled list makes equipment an autodialer. The Third, Seventh, and Eleventh Circuits now align with the Supreme Court's interpretation, after the Second, Sixth, and Ninth Circuits had previously taken broader views.

One more clarification matters for anyone running outbound campaigns. The FCC's 2020 Declaratory Ruling confirmed that call volume is not probative of autodialer status. The only test is whether the equipment can dial random or sequential numbers without human intervention. Running a high-volume campaign against a permissioned list does not, by itself, trigger autodialer liability.

For a managed service like My AI Call Center, this distinction shapes how campaigns are built. Calls run against approved, permissioned, or reviewed contact lists, with list source and consent records checked before launch, rather than indiscriminate dialing into randomly generated numbers.

One caution before you assume the autodialer question settles everything: the TCPA's separate prohibition on artificial or prerecorded voices still applies. The FCC has confirmed that AI-generated voices fall under that provision, requiring prior express consent even when the dialing itself never touches the autodialer definition. Compliance means getting both pieces right, and requirements vary by location, industry, and consent status, so appropriate legal guidance before launch is worth the effort.

The Compliance Trap Most Businesses Miss: AI Voices Are a Separate Rule

Many businesses assume that clearing the autodialer hurdle means they're fully compliant. It doesn't. Even if your dialing technology passes the Supreme Court's narrow test, a second, separate TCPA rule governs the voice on the call itself, and that rule now explicitly covers AI-generated voices.

The autodialer question was largely settled in 2021. In Facebook, Inc. v. Duguid, the Supreme Court ruled unanimously, 9-0, that a device qualifies as an autodialer only if it can store or produce numbers using a random or sequential number generator. Dialing from a customer-supplied list of pre-compiled numbers doesn't qualify, even when the system dials those numbers automatically. The Third Circuit reaffirmed this in 2024, calling it a "common-sense" reading of the law.

For businesses running structured campaigns against permissioned lists, that's genuinely good news. The FCC's 2020 Declaratory Ruling added another clarification: the volume of calls you send is not probative of autodialer status. Whether you call fifty numbers or fifty thousand, the test is only whether the equipment dials random or sequential numbers without human intervention. High volume alone doesn't create liability.

Here's the part that catches businesses off guard. The TCPA contains a separate prohibition on calls using an "artificial or prerecorded voice," and in February 2023, the FCC issued a Declaratory Ruling confirming that AI-generated voices fall squarely under that provision. This means prior express consent is required for the voice itself, completely independent of whether your dialing equipment qualifies as an autodialer.

So the compliance picture is actually two questions, not one:

• Is your dialing method an autodialer? Under the Duguid standard, dialing an approved, permissioned list generally is not.
• Does your call use an artificial or prerecorded voice? If it's AI-generated, yes, and prior express consent is required.
• Do you have consent records that prove it? Consent for the voice must exist before the call is placed, which makes your list source and permission documentation the primary safeguard.

The stakes are real. TCPA statutory damages run from $500 to $1,500 per call, which is why class action litigation has followed this law for years. A business that clears the autodialer test but dials an AI voice without proper consent is still exposed.

This dual standard is why list discipline matters more than dialing technology. At My AI Call Center, every campaign starts with a review of the list source and consent records before anything launches, and bought lists without clear permission records are flagged and usually declined. The consent for the AI voice has to be on file before the first call, not discovered afterward.

The practical takeaway: passing one compliance test tells you almost nothing about the other. Treat your dialing method and your voice technology as two separate boxes to check, and get consent documentation for both before any campaign goes live.

How to Run High-Volume Campaigns Without Triggering Autodialer Liability

Most businesses assume that calling a lot of people is what gets them in legal trouble. The federal rules say otherwise. Under the Telephone Consumer Protection Act, what matters is how the numbers were generated, not how many calls you place.

The Supreme Court settled this in Facebook, Inc. v. Duguid (2021), a unanimous 9-0 decision holding that an autodialer is only equipment that stores or produces numbers using a random or sequential number generator. Dialing from a list you already compiled doesn't qualify, even if the dialing itself is automated. The Third Circuit reaffirmed this in 2024, calling it a "common-sense" rule: if the customer supplied the set of phone numbers, the system isn't an autodialer.

The FCC went further in its 2020 Declaratory Ruling, clarifying that the volume of calls sent is not probative of autodialer status. That ruling matters if you're running a database reactivation blitz or speed-to-lead follow-ups where new leads get called within minutes. High volume, on its own, doesn't create autodialer liability.

So what actually protects you? The list.

• Use approved, permissioned, or reviewed contact lists, never numbers generated by the system.
• Check consent records and list source before launch. Bought lists without clear permission records should be declined.
• Document that every number came from a real customer relationship, form submission, or prior interaction.
• Honor opt-outs immediately and carry DNC requests across all campaigns.

This is the discipline My AI Call Center applies to every campaign: list source and consent records get reviewed before anything launches, and if the list won't support the campaign, we say so before you spend anything.

Here's the part that trips people up. Even if your dialing is compliant, the TCPA's separate "artificial or prerecorded voice" provision still applies. The FCC's February 2023 Declaratory Ruling confirmed that AI-generated voices are subject to TCPA restrictions, which means you need prior express consent for the voice itself. Before any AI calling campaign launches, document that consent, disclose that the call is AI-assisted, and give recipients a way to opt out or request a human.

One caution on geography. Before 2021, the Second, Sixth, and Ninth Circuits read the autodialer definition broadly, and the Ninth Circuit's earlier holding conflicted with the FCC's 2020 clarification. Duguid is the controlling precedent now, but litigation strategies in places like California sometimes reference older, broader interpretations. That's why campaign requirements vary by location, industry, and consent status, and why you should get local legal guidance before launch. A free campaign review can surface the consent and list questions worth raising with your attorney.