The TCPA ruling about cell phones your compliance team will misread

A district court in Missouri just gave the telemarketing defense bar a nice gift, and within a week I expect to hear it quoted in conference rooms as "cell phone DNC claims are dead." That reading is wrong, and acting on it is how outbound programs get hurt. The ruling is real, it is interesting, and it is narrow: one judge, one district, one statute prong, no circuit court behind it. Here is what actually happened in the case, what it changes, and the parts of your exposure it does not touch at all.

What actually happened

The case is Koeller v. OX AppSec Security, Inc., out of the Eastern District of Missouri, decided September 30, 2026. The plaintiff alleged that OX called him in violation of the Telephone Consumer Protection Act's Do-Not-Call provisions. The wrinkle: he alleged the calls went to his cell phone, not a residential landline.

The court dismissed the claim. Its holding, in one sentence: the TCPA's private right of action for Do-Not-Call violations, 47 U.S.C. § 227(c)(5), does not extend to cellular telephone subscribers, so a plaintiff whose only number is a cell phone has no cognizable claim under that provision. The full writeup from the firm that reported it is worth reading directly: Greenspoon Marder's case note.

If your first reaction is "wait, the TCPA is mostly about cell phones," you are thinking of a different subsection. That reaction is the whole reason this case matters, so it is worth walking the statute once, slowly.

The two prongs people constantly blend together

The TCPA is not one claim. It is several, and they live in different subsections with different plaintiffs, different defenses, and different damages.

The Do-Not-Call prong, § 227(c), is the registry prong. Congress directed the FCC to protect the privacy of "residential telephone subscribers" from solicitation calls, the FCC built the national registry on top of that language, and § 227(c)(5) gives a subscriber who gets more than one violating call from the same entity in twelve months a private action. Damages follow the statute's familiar ladder: $500 per violation, trebled to $1,500 when the violation is willful or knowing.

The cell phone prong, § 227(b), is the one that funds the plaintiff's bar. It bars calls made with an autodialer, an artificial or prerecorded voice, or both to a cell number without the called party's prior express consent, and § 227(b)(3) gives the called party a private action at $500 to $1,500 per call. Multiply that by a campaign that dialed a list for six weeks and you understand why TCPA class settlements have commas in them.

Koeller was a (c)(5) case. The plaintiff sued under the registry prong over calls to his cell. The court read "residential telephone subscriber" to mean what it said in 1991: telephone service at a residence. Claim dismissed.

The reasoning, in plain language

The court's logic runs five steps, and the first step is the one with the long shadow.

First, it applied Loper Bright Enterprises v. Raimondo, the 2024 Supreme Court decision that retired Chevron deference. Under Loper Bright, courts construe statutes themselves instead of deferring to an agency's reading. So the court did not ask what the FCC thinks "residential telephone subscriber" covers. It asked what Congress wrote.

Second, plain meaning: "residential telephone subscriber" is telephone service connected to a residence.

Third, context: in 1991, when Congress passed the TCPA, residential service and cellular service were distinct categories, priced differently, owned by different kinds of customers, and treated separately throughout the statute.

Fourth, the drafting tell: Congress expressly referenced cellular telephone service in other parts of § 227 and did not do so in § 227(c). Where Congress included cell phones in one subsection and left them out of another, courts read the omission as deliberate.

Fifth, the limit on agency power: the FCC's 2003 order, and a run of courts after it, treated wireless subscribers as covered by the DNC rules. This court said an agency cannot expand a private right of action beyond what Congress authorized. The registry rules themselves still stand. The private lawsuit is what got narrowed.

That last distinction, rules versus private action, is where most of the hot takes go off the rails. The court did not deregulate calling cell phones. It did not strike the Do-Not-Call registry. It did not touch § 227(b). It held that one private cause of action, in one district, belongs to residential-line subscribers.

Why 1991 keeps deciding cases in 2026

The statute's age is not trivia here. It is the engine of the ruling, and understanding it tells you which way the argument travels.

When Congress wrote the TCPA in 1991, a cellular subscriber paid for incoming airtime by the minute. Congress protected cell users in § 227(b) with a consent model, partly because an unsolicited call cost the recipient actual money. Residential lines got the privacy model in § 227(c): a registry, enforced by rules, with a private action for the subscriber. The two subsections were built for two different problems, cost intrusion and home intrusion, on two different network categories that barely overlapped.

Then the networks converged. Most American households are now wireless-only, the FCC's 2003 order folded wireless subscribers into the DNC protections, and two decades of district courts followed along without anyone pressing hard on the 1991 drafting. What Loper Bright changed is the pressure. Courts now owe the statute their own reading, and the statute says "residential telephone subscriber." The Missouri court read those three words and stopped. Other courts will read the same words with the FCC's thirty years of practice in the other pan of the scale. That disagreement is exactly what circuit review exists to settle, and until it settles, the map is patchy by district.

For an operator, the lesson is that the statute you comply with was drafted for a network that no longer exists. The categories in your dial plan (cell, landline, VoIP did not exist as a category at all in 1991) do not map cleanly onto the categories in the law. Where the mapping is contested, your protection is not a clever reading. It is consent, scrubbing, and records, which survive any mapping.

What the ruling does not change

Run the list, because every item on it is a live exposure for an outbound program.

Section 227(b) claims are untouched

The autodialer and prerecorded-voice claims, the consent claims, the revocation claims: all of § 227(b)(3) is exactly what it was the day before the ruling. A cell-only plaintiff who alleges prerecorded calls without consent still has the same federal claim in the same courthouse that just dismissed Koeller. If your cell campaigns rest on consent you cannot document, this case does not help you at all.

One operator-relevant nuance from the 2021 Supreme Court decision in Facebook v. Duguid: the autodialer element now requires a random or sequential number generator, so list-based dialing fights those claims on the consent and prerecorded-voice elements rather than on the dialer hardware. The exposure moved. It did not evaporate.

The Do-Not-Call rules still apply

The registry regime is an FCC rule set enforced by the FCC, the FTC where the Telemarketing Sales Rule overlaps, and state attorneys general. Koeller narrowed who can bring a private lawsuit under one prong. It narrowed nothing about what the regulators can do to you. Stop scrubbing against the registry and your problem arrives with a subpoena instead of a complaint, which is not an improvement.

State mini-TCPAs still apply

Florida, Oklahoma, Washington, and a growing list of states run their own telephone-solicitation statutes with their own per-call damages and their own definitions, several of which reach conduct the federal TCPA does not. Nothing in a Missouri federal court's reading of a federal subsection touches any of them.

Your carrier and the analytics layers still apply

The private right of action was never the only enforcement mechanism for call behavior that consumers hate. Carrier spam scoring, STIR/SHAKEN attestation, and the platforms' complaint feedback loops penalize the same behavior the statutes do, faster and without a docket number. A campaign that burns numbers gets filtered whether or not anyone sues.

The exposure map after Koeller

For a single-page view of where things stand, this is the honest table.

Claim prong	Source	Who can pursue it after Koeller	Typical damages	Effect of the ruling
Do-Not-Call registry violations	47 U.S.C. § 227(c)(5)	Residential-line subscribers, per this court; cell-only plaintiffs dismissed in E.D. Mo.	$500 per violation, up to $1,500 willful	Narrowed, one district, no circuit ruling
Autodialer / prerecorded voice to cells without consent	47 U.S.C. § 227(b)(3)	The called party	$500 per call, up to $1,500 willful	None
Consent revocation honored late or never	§ 227(b) via FCC rules	The called party	Same as above	None
DNC and TSR rules, regulator-enforced	FCC / FTC / state AGs	The government	Forfeitures, consent orders, injunctions	None; rules stand
State telephone-solicitation statutes	State law (FL, OK, WA, others)	Varies by state	Per-call, often uncapped by the federal ladder	None

Read the middle column of the first row carefully. "Cell-only plaintiffs dismissed in E.D. Mo." is not "cell-only plaintiffs dismissed." A doctrine with no circuit endorsement is a litigation position, not a safe harbor.

Where this actually touches your dialer

Here is the part that connects to the work we do. Koeller is a case about the call being placed. The exposures that live on your dialer are mostly about what happens after the call is answered, and those did not move an inch.

The abandonment math did not move

The FTC's Telemarketing Sales Rule caps abandoned calls at 3 percent of answered calls per campaign, measured over 30 days. The measurement trap is unchanged and it is still the one I see most often: when your answering machine detection hangs up on a live person because it scored them as a machine, that call never enters your abandoned-call count at all. The person experienced a dropped call. Your dashboard never recorded one. That is not a loophole, it is a liability wearing a costume, and I wrote the full mechanics in the 3% rule and how AMD accuracy affects it. The denominator games are covered in detail in how predictive dialer abandonment rate actually gets computed.

The evidence requirement did not move

Every claim in that table, and every regulator request, eventually asks what happened on one specific call at one specific timestamp. Population statistics do not answer that question. A per-call detection record does: verdict, cause, latency, and a pointer to retained audio. This is the same argument I made about detection logs as defense exhibits, and a ruling about who can sue does not change what you will wish you had logged. Detection supports compliance. It is not compliance, and it substitutes for neither your counsel nor your written policy.

What it does change, narrowly

Inside the courtroom, the ruling has real effects, and they are worth knowing even though none of them reach your dial plan.

Cell-only DNC cases now settle for less. A claim that survives a motion to dismiss settles for multiples of one that does not. Defense counsel in the Eighth Circuit now have a documented basis to attack the (c)(5) prong on cell-only pleadings, and plaintiffs' counsel price that risk into demand letters.

Pleading patterns shift. Expect plaintiffs to respond the way they always do: plead § 227(b) counts alongside the DNC count, allege prerecorded voice or consent violations that survive Duguid, and keep the case alive on the prong the Missouri court never touched. A complaint that pleads around a dismissal ruling has not lost anything.

Forum selection gets sharper. Plaintiffs choose courthouses for a living, and a pure-DNC claim now looks worse in the Eastern District of Missouri than it did in September. Expect fewer of those filings there, and expect defendants to remove toward it where they can.

None of this is nothing. It is also none of it a reason to change one setting on your dialer.

A three-question test for TCPA headlines

You will see more of these headlines as the post-Loper Bright wave works through the district courts, in both directions. Before a headline changes anyone's behavior, run three questions on it.

Which subsection? "TCPA" in a headline tells you nothing. (b) and (c) are different statutes in practice, with different plaintiffs and different defenses. A ruling about (c)(5) is not a ruling about robocall consent no matter how the headline is written.

Which court? A district court decision binds nobody outside that district, and it is a data point even inside it. A circuit court decision changes doctrine. The gulf between those two is where bad compliance decisions get made.

Who was the plaintiff? Koeller was a cell-only subscriber suing under the residential prong. Half the rulings that get shared as "TCPA is dying" are one unusual plaintiff testing one edge of one subsection, which is how doctrine gets built, not how liability ends.

Run the test on this case: subsection (c)(5), one district court, a cell-only plaintiff on the residential prong. Interesting, quotable, and a reason to change exactly nothing before the circuit courts speak.

What to do Monday morning

Short list, in order of how often I see it skipped.

• Keep scrubbing every list against the national registry and your internal DNC list, cell numbers included. The rules did not change, and the regulators who enforce them did not read this case.
• Audit consent records per campaign, per list source. Consent provenance is the whole (b) defense.
• Handle revocation the day it arrives, on every channel, and write down where you recorded it.
• Recheck your abandonment measurement against the 3 percent rule using answered-call denominators that include AMD-disconnected humans. If the number only looks good because misclassified people vanish, it is not good.
• Pull 50 recent calls and score detection verdicts against what actually answered. That is the calibration loop, and it is the only honest way to know your misroute rate.
• Retain the per-call detection log with timestamps and latency, under a retention policy your counsel signed off on.

None of that is new advice. That is the point. Nothing in Koeller retired any of it.

Where this goes next

The question the Missouri court answered is now open at the circuit level, and neither side's lawyers will treat one district court as the answer. Expect copycat motions to dismiss in cell-only DNC cases, expect pushback citing the FCC's 2003 order and the pre-Loper Bright case law, and expect the first circuit panel to settle it eventually. When the Eighth Circuit or any other circuit takes it up, the calculus changes for real. Until then, this is a defense argument that sometimes works, in one district, on one prong.

The strategic takeaway for an outbound shop is unglamorous. The cheapest time to handle TCPA exposure is before the list is dialed: consent you can prove, scrubbing you can show, abandonment you measure honestly, and per-call records you can produce. A ruling that narrows one private right of action changes none of that arithmetic. It just changes who gets to be the plaintiff on one particular claim, in one courthouse, for now.

This article is educational, not legal advice. Decisions about dialer configuration, list handling, and compliance policy belong with qualified counsel who has read your actual program.