The Telephone Consumer Protection Act, at 47 U.S.C. § 227, is unusual among consumer statutes: it hands ordinary people a private claim with fixed statutory damages per message, and it does not require you to prove you lost a dollar. That is why unwanted calls and texts generate so much litigation, and why businesses that send marketing messages treat consent records as compliance infrastructure.
The central rule is simple to state. Marketing calls and texts placed to a mobile number using an automatic telephone dialing system, or using an artificial or prerecorded voice, generally require the recipient's prior express written consent. Non-marketing informational calls made with the same technology require prior express consent, which need not be written. Consent can be revoked, and once revoked the sender must stop.
What "prior express written consent" actually requires
The written consent standard is specific, and box-checking alone does not satisfy it. The agreement must be signed — an electronic signature counts — and must clearly authorize the identified seller to deliver marketing messages using an autodialer or prerecorded voice to a number the person designates. It must disclose that consent is not a condition of purchase. A pre-checked box, or consent buried in a terms-of-service document the consumer never saw, is where most defense arguments fail.
Two practical consequences follow. For consumers: giving a store your mobile number at checkout is not, by itself, written consent to marketing texts. For businesses: the consent record has to capture the exact disclosure language, the timestamp, the IP address or channel, and the number consented to, because the burden of proving consent sits with the sender.
Watch out: Consent given to one company does not transfer to its marketing partners or to companies that later buy its lead list. Rules on how consent may be shared across sellers have been revised and litigated repeatedly; as of mid-2026, confirm the current requirements at fcc.gov rather than relying on older summaries.
The autodialer question
Whether equipment qualifies as an automatic telephone dialing system has been the most litigated issue under the statute. In Facebook, Inc. v. Duguid (2021), the Supreme Court read the statutory definition narrowly, holding that the device must have the capacity to store or produce telephone numbers using a random or sequential number generator. Systems that dial from a curated customer list, without that generator capability, fall outside the definition.
That decision did not end TCPA litigation; it redirected it. Claims now cluster around three theories that do not depend on the autodialer definition at all: artificial or prerecorded voice calls, which are separately restricted; Do Not Call violations, which turn on the fact of telemarketing rather than the technology; and revocation failures, where the sender kept messaging after being told to stop.
Do Not Call rights, layered on top
Two separate protections operate alongside the consent rules. The national Do Not Call Registry, maintained under FTC and FCC authority, lets you register a residential or wireless number free of charge and permanently. Telemarketers generally may not call a registered number, subject to exceptions for existing business relationships, prior written permission, tax-exempt nonprofit organizations, and calls that are not solicitations. Receiving more than one telemarketing call within a twelve-month period on a registered number supports a private claim under the statute's Do Not Call provisions.
Separately, every seller and telemarketer must maintain an internal do-not-call list and honor a request to stop calling. That internal request is often stronger than registry status, because it survives an existing business relationship. The FTC's Telemarketing Sales Rule adds requirements on disclosures, calling hours, abandoned calls, and caller ID transmission; its current text is in the FTC rules library.
| Message type | Typical requirement | Main exception |
|---|---|---|
| Marketing text or call to a mobile, autodialed or prerecorded | Prior express written consent | None that is broad; consent must be provable |
| Informational call or text, autodialed or prerecorded | Prior express consent | Often satisfied by providing the number for that purpose |
| Prerecorded telemarketing to a residential line | Prior express written consent | Limited exemptions, including some nonprofit calls |
| Live telemarketing call to a registered number | Do Not Call rules apply | Existing business relationship; written permission |
| Any call after you say stop | Must cease | Narrow; internal do-not-call request controls |
Revoking consent so it sticks
Consent is not permanent. A consumer may revoke consent to autodialed and prerecorded calls and texts through any reasonable method, and the FCC has adopted rules confirming that senders must treat common stop words — including "stop," "quit," "end," "revoke," "opt out," and "unsubscribe" — as effective, and must honor revocation within a short defined period rather than at their convenience. Rules in this area have been adjusted and their effective dates extended more than once; as of mid-2026, check the current text at fcc.gov before relying on a specific deadline.
- Reply with a plain stop word. Use the exact word the message offers if one is given, and also reply "STOP" to be safe. Screenshot the reply and the confirmation.
- Send a second revocation through another channel. Email the company's published contact address stating that you revoke consent for all calls and texts to your number and asking to be added to its internal do-not-call list.
- Log everything that arrives afterward. Date, time, number, content, and whether it was a call or text. Screenshots with visible timestamps are the core evidence.
- Register the number. Registry status strengthens a Do Not Call theory and costs nothing.
- Identify the actual sender. Displayed caller ID is frequently spoofed. The seller whose product is advertised is the party that matters, and is often liable for a marketing vendor's conduct.
Damages, and why records decide value
The statute provides statutory damages of $500 for each violating call or message, which a court may treble to $1,500 where the violation was willful or knowing, in addition to injunctive relief. Because damages are per message, a campaign that sent dozens of texts becomes a meaningful claim even without out-of-pocket loss.
Proof is the constraint. Courts want to see the messages themselves, the number they came to, and evidence that consent was absent or revoked. A phone that has been wiped, or a screenshot without a visible date, weakens an otherwise strong case. Preserve the device records and export the message thread before you change phones.
Practical step: Report unwanted calls to the FCC and the FTC even if you never sue. Complaint data drives enforcement actions and traceback efforts against the carriers routing illegal traffic, and the complaint itself is a dated record of what you received.
Many TCPA cases are filed as class actions, since a single campaign hits thousands of numbers identically. If you receive a class notice about calls or texts you recognize, the options are set out in our guide to class action notices and opting out. Consumer contracts increasingly try to route these claims to individual arbitration — a dynamic explained in arbitration clauses and class-action waivers.
Quick answers
A collector keeps calling my cell. Is that a TCPA claim?
It can be, if the calls use a prerecorded or artificial voice, or an autodialer as narrowly defined, and you either never consented or revoked consent. Debt collection calls are informational rather than marketing, so the written-consent standard usually does not apply — but revocation does. Collection conduct is separately regulated, and our guide to FDCPA debt collection rights covers the call-frequency and cease-communication rules.
I gave a company my number years ago. Did I consent forever?
No. Consent tied to a specific purpose does not automatically extend to marketing, and any consent can be revoked at any time by any reasonable means. Businesses sometimes argue that an old account form covers current campaigns; the answer depends on what the form said and whether it met the written-consent disclosure requirements at the time.
Does registering with the Do Not Call list stop scam calls?
Not in practice. Registration binds legitimate telemarketers, who face liability for calling registered numbers. Outright fraudsters ignore it, spoof caller ID, and often operate outside the United States, which is why the enforcement focus has shifted toward the carriers and gateway providers that route the traffic. Registration is still worth doing, both to reduce lawful calls and to support a claim.
Can a business be liable for calls made by a vendor it hired?
Often yes. Sellers can be held responsible for calls made on their behalf under agency principles, including apparent authority and ratification. "Our lead generator did it" is not a reliable defense, which is why compliant businesses audit vendor consent records, require indemnities, and keep the underlying consent evidence themselves rather than trusting a supplier to have it.
A sensible order of operations
Revoke clearly and in writing, keep the screenshots, register the number, and log every message that arrives afterward with a visible timestamp. Identify the seller behind the campaign rather than the spoofed number. File complaints with the FCC and FTC while the record is fresh. If the messages continued after a documented revocation, the per-message damages structure means the file you built in the first month usually determines what the claim is worth. Related consumer guides sit in the consumer rights and civil claims pathway, and a modest single-sender claim may fit in small claims court.