The FCC’s 2015 robocall order exempted free fraud and care alerts and restated that callers must prove consent
The Declaratory Ruling and Order released July 10, 2015 exempted conditioned bank and healthcare alerts, allowed revocation by any reasonable means and gave one call on reassigned numbers. In our reading, it made consent a record each caller had to keep.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
The Declaratory Ruling and Order released July 10, 2015 exempted conditioned bank and healthcare alerts, allowed revocation by any reasonable means and gave one call on reassigned numbers. In our reading, it made consent a record each caller had to keep.
- The week On July 10, 2015, the Federal Communications Commission released a Declaratory Ruling and Order under the Telephone Consumer Protection Act.
- The free-message exemptions: what could be sent without consent The exemptions rested on section 227(b)(2)(C) of the statute.
- Consent and revocation: how consent began and ended For healthcare, it applied that view to HIPAA’s own categories.
- Reassigned numbers: one call, then the caller’s risk The order then gave callers a limited allowance.
- The equipment question: which dialers the rules reached The statute defines an autodialer by capacity.
- How the order reached banks, health systems and agencies The order described how institutions identify events that trigger fraud alerts.
- Objections, and what the record shows The record says otherwise.
- Checks a compliance or contact-center leader could run These checks are our suggestions, not requirements of the 2015 order or the current rules.
- With hindsight The court set aside two other parts.
The week
On July 10, 2015, the Federal Communications Commission released a Declaratory Ruling and Order under the Telephone Consumer Protection Act. The Commission had adopted it on June 18, 2015. [1] The document carries the number FCC 15-72 and is filed in CG Docket No. 02-278. [1]
The vote was split. The order records separate statements from Chairman Wheeler and Commissioner Clyburn, partial dissents from Commissioners Rosenworcel and O’Rielly, and a dissent from Commissioner Pai. [1]
The order opened with the scale of the problem. It said that, month after month, unwanted robocalls and texts top the list of consumer complaints the Commission receives. [1] It addressed 19 petitions and a letter from the National Association of Attorneys General, and it declined a petition for rulemaking. [1] It described the whole as resolving 21 separate requests for clarification or other action. [1]
The petitioners included the American Bankers Association, the Consumer Bankers Association, the American Association of Healthcare Administrative Management and United Healthcare Services. [1] Retailers, messaging companies and a consumer lender filed as well. [1] The order dealt with them by issue rather than one at a time. [1]
In our reading, four rulings mattered most to banks, health systems and their service providers.
The first concerned dialing equipment. The Commission reaffirmed that equipment generally has the capacity to store or produce, and dial, random or sequential numbers even when it is not presently used for that purpose. [1] That included equipment used to call a set list of consumers. [1] It reiterated that predictive dialers meet the statutory definition. [1]
The second concerned consent. The Commission clarified that consumers may revoke consent through any reasonable means. [1] It denied a request that callers be allowed to designate the exclusive means of revocation. [1] It said consumers may revoke in any manner that clearly expresses a desire not to receive further messages. [1]
The third concerned reassigned wireless numbers. The Commission said the statute requires the consent of the current subscriber or customary user, not the intended recipient. [1] It allowed callers one additional call after reassignment, where the caller had no knowledge of it and a reasonable basis to believe it had consent. [1] If that call did not reveal the reassignment, the order deemed the caller to have constructive knowledge of it. [1]
The fourth concerned exemptions. Using its statutory power to exempt calls to wireless numbers that are not charged to the consumer, the Commission exempted certain free messages from financial institutions and healthcare providers. [1] The bank messages covered fraud and identity theft, data security breaches, steps to remedy a breach and money transfers. [1] The healthcare messages covered a list of calls with a healthcare treatment purpose, which also had to comply with HIPAA privacy rules. [1] Each exemption came with seven conditions. [1]
The order also clarified a consent question for healthcare. It said giving a phone number to a healthcare provider is prior express consent for HIPAA-subject healthcare calls by a covered entity and its business associates. [1] The consent reaches calls within its scope, absent instructions to the contrary. [1]
Other rulings reached app makers, collect-call services and on-demand texts. [1] The order confirmed that text messages are calls under the statute. [1] It said nothing in the Communications Act or the Commission’s rules prohibits carriers and VoIP providers from offering consumer-initiated call blocking. [1]
For an institution that sends account alerts, appointment reminders or collection calls, the week’s document was a rulebook for its outbound contact operation. The rest of this piece is about what that rulebook asked an institution to keep.
The FCC’s July 2015 robocall order is remembered for the fights it started. A court challenge to its rulings on autodialers and reassigned numbers was decided in 2018. [2] For a chief compliance officer at a bank or a director of patient access at a health system, the order did something more immediate. It turned every automated call into a question about records.
Our argument is that the order confirmed consent as something the caller had to prove, and added the identity of the person holding a number to that list. It restated, as longstanding Commission precedent, that if any question arises about consent, the burden is on the caller to prove it obtained the necessary prior express consent. [1] It said the evidentiary value of business records gave callers reasonable ways to carry that burden, and that it expected responsible callers to keep proper business records tracking consent. [1] It also put the burden on the caller to show a reasonable basis for believing it had consent, and that it lacked knowledge of a reassignment. [1] It let consumers revoke by any reasonable means. [1] And its new exemptions for free fraud alerts and care messages came with conditions that, in our reading, an institution could meet only with good data about each message it sent. [1]
That framing places the work with the people who run contact centers, customer data and messaging vendors, alongside the lawyers. It also explains why the order’s later history matters. Parts of it were set aside in 2018, and the Commission changed the revocation rules in 2026. [2] [5]
We take the order in four parts:
- The free-message exemptions: what banks and providers could send without consent, and on what terms.
- Consent and revocation: how consent was created, and how it ended.
- Reassigned numbers: who the “called party” was, and how much a caller had to know.
- The equipment question: which dialing systems the rules reached.
The parts do not overlap. The first covers calls that needed no consent, the second the life of consent where it was needed, the third whose consent counted, and the fourth which calls were regulated at all. After them we look at how the order reached different institutions, take the objections in order and end with checks a compliance or contact-center leader could run.
The free-message exemptions: what could be sent without consent
The statutory door
The exemptions rested on section 227(b)(2)(C) of the statute. It lets the Commission exempt calls to wireless numbers that are not charged to the called party, subject to conditions it prescribes in the interest of privacy. [1] The Commission read the “no charge” requirement to exclude notifications that count against a recipient’s plan minutes or texts. [1]
That reading put a commercial question at the front of the exemption. An institution could rely on it only if its messages reached customers free of charge. [1] The order recorded the banking petitioner’s statement that institutions would work with carriers and third-party providers to make sure of that. [1] In our reading, “free to the end user” was an operational fact to arrange and evidence, not a legal label to assert.
What banks could send
The banking petition asked for four categories of messages. [1] The Commission granted all four. The first two covered transactions and events suggesting a risk of fraud or identity theft, and possible breaches of the security of customers’ personal information. [1] The other two covered steps consumers can take to prevent or remedy harm from a breach, and actions needed to receive pending money transfers. [1]
The order explained its reasoning for the first three as exigency. It said a quick, timely communication could prevent considerable consumer harm, or help mitigate it, and that requiring prior consent could make effective communication impossible. [1] It rejected the argument that banks could simply ask customers to opt in. [1] For money transfers, it noted that recipients may have no relationship with the sending institution. [1]
The order drew a firm line around content. Breach and remediation calls could not carry telemarketing, cross-marketing, solicitation, debt collection or advertising of any kind. [1] The Commission said informing a consumer of identity theft is distinct from marketing products the consumer may use to prevent or fix it. [1]
Seven conditions for banks
The order set seven conditions for each exempted bank call or text. [1]
- Messages go only to the wireless number provided by the customer of the financial institution.
- Each states the institution’s name and contact information, at the start of a voice call.
- Each is limited to the exempted purposes, with no telemarketing, cross-marketing, solicitation, debt collection or advertising content.
- Each is concise: generally one minute or less for voice and 160 characters or less for text.
- No more than three messages per event over a three-day period for an affected account.
- Each offers an easy opt-out: a key-press or voice mechanism on live calls, a toll-free number on voicemail, and a reply of “STOP” on texts.
- Opt-out requests are honored immediately.
The Commission explained the frequency limit. It declined the request for no limit, saying that would diminish consumers’ privacy rights, which the statute required it to consider. [1] It said the limit applied per event, per affected account, per three-day period. [1]
It also accepted a narrower opt-out. It agreed with the banking petitioner that opting out of one category should not opt a consumer out of all financial calls for that account, and let institutions tailor the opt-out notice to say that the choice applied to one of the four categories. [1] That choice had a cost. In our reading, an institution relying on it needed to record each opt-out by category and account, and route every later message through that record.
What providers could send
The healthcare petition sought a broader exemption than it received. It listed calls ranging from appointment reminders to payment notifications and Social Security disability eligibility. [1] The Commission said it saw exigency in calls such as post-discharge follow-up and prescription notifications, but not in account and payment calls. [1]
It granted the exemption only for calls with a healthcare treatment purpose. [1] The list was appointment and exam confirmations and reminders, wellness checkups, hospital pre-registration instructions, pre-operative instructions and lab results. [1] It also covered post-discharge follow-up intended to prevent readmission, prescription notifications and home healthcare instructions. [1] It excluded calls with telemarketing, solicitation or advertising, and calls with accounting, billing, debt-collection or other financial content. [1]
Seven conditions for providers
The healthcare conditions followed the bank model with tighter limits. [1] Messages could go only to the wireless number provided by the patient. They had to state the provider’s name and contact information, stay within the permitted purposes and comply with HIPAA privacy rules. [1] Voice calls were generally one minute or less and texts 160 characters or less. [1]
Frequency was capped at one message per day, up to three voice calls or texts combined per week from a specific healthcare provider. [1] The opt-out mechanics matched the bank conditions, and opt-outs had to be honored immediately. [1]
The order added a privacy caution of its own. It said HIPAA privacy rules control the content where a message would relate sensitive or personal information. [1] In our reading, that tied the length limit to a content design problem. A useful 160-character lab-result text had to say enough to prompt action and too little to harm the patient if someone else saw the phone.
“STOP” as the only door for exempt texts
One condition ran the other way from the order’s revocation ruling. For exempted texts, a reply of “STOP” was to be the exclusive means by which consumers could opt out of such messages. [1] The same order denied callers the right to designate an exclusive revocation method for consent-based calls. [1]
The difference, as we read it, followed from the legal basis. Consent-based calls rest on the consumer’s permission, which the Commission said the consumer may withdraw in any reasonable way. [1] Exempted calls rest on the Commission’s own conditions, and it could define the opt-out as one of them. [1] An institution sending both kinds of text to the same customer had to treat a “STOP” differently from a free-form reply such as “please stop texting me,” and still honor both.
Evidence for each exempted message
An exemption is a defense, and a defense needs evidence. The order said the exemption applies only if messages are not charged to the recipient and the institution complies with the conditions. [1] For any disputed message, an institution relying on it would want to show the number’s source and the event that triggered the message. It would also want the template used, the count for that event and the opt-out status when the message was sent.
That evidence can live in different places: fraud events in a fraud platform, numbers in a customer system, templates with a messaging vendor and opt-outs in another store. The conditions joined them. A compliance team that could not join them on demand could not show the exemption applied.
What the exemptions asked of an institution
Read together, the conditions described a message system with specific properties. It knew which number each customer or patient had provided. It knew why each message was sent and could show the content contained nothing promotional. It counted messages by event and by week. It captured opt-outs in every channel and applied them at once.
The order made those properties conditions of a legal exemption. [1] In our reading, that changed who in an institution owned them. A message template became a compliance artifact, and a frequency counter became evidence.
Consent and revocation: how consent began and ended
Consent from a phone number
Outside the exemptions, autodialed and prerecorded calls to wireless numbers still needed prior express consent. [1] The order restated the Commission’s long-standing view that people who knowingly release their phone numbers have in effect invited calls at that number, absent instructions to the contrary. [1]
For healthcare, it applied that view to HIPAA’s own categories. Giving a number to a healthcare provider counted as consent for HIPAA-subject healthcare calls by a covered entity and its business associates. [1] The clarification reached only calls within the scope of the consent given, and only absent instructions to the contrary. [1] The Commission said its clarification extended only to calls subject to HIPAA. [1]
The order gave an example of scope. A number given on admission for scheduled surgery would support calls about that surgery or follow-up procedures. [1] In our reading, that made the purpose of collection part of the consent record. A registration system that stored a number without the context in which it was given would leave the scope question unanswerable later.
Consent during incapacity
The order addressed a problem particular to hospitals. Where a patient cannot consent because of medical incapacity, prior express consent for HIPAA-subject healthcare calls may come from a third party. [1] That consent lasts only for the period of incapacity, after which the caller must obtain the patient’s own consent. [1]
The order said incapacity was a legal concept, assessed with reference to legal indicia rather than in a loose colloquial sense. [1] In our reading, that required a hospital to tie the source of a phone number to a status that changes. A number provided by a family member at admission might support calls during an admission and not after discharge.
Revocation by any reasonable means
The revocation ruling answered a petition from a consumer lender. It had asked the Commission to rule that consent could not be revoked, or that callers could designate the exclusive methods of revocation. [1]
The Commission refused both. It said consumers have a right to revoke, using any reasonable method including orally or in writing. [1] It gave examples: a consumer-initiated call, a response to a call from the caller, or an in-store bill payment location. [1] It said “callers typically will not find it overly burdensome to implement mechanisms to record and effectuate a consumer’s request to revoke his or her consent.” [1]
The Commission compared the ruling to its existing rules. Prerecorded telemarketing calls already had to offer an automated opt-out, and an earlier exemption had required a reply of “STOP” for texts. [1] It said consumers must be able to respond to an unwanted call in a way that ends further calls. [1]
What revocation asked of an institution
A bank or health system can hear from the same person through several doors. There are branch staff and front desks, call centers, collection vendors, patient portals, chat and mail. Under the 2015 ruling, a clear request through any reasonable door could end consent. [1]
In our reading, the practical consequence was a data problem. A revocation heard by a branch teller had to reach the dialer the collections vendor used. A “stop” said to a scheduling agent had to reach the reminder platform. The institution needed one record of consent status per number, fed from every channel and read by every sending system.
The exempt bank and healthcare messages carried their own rule: opt-outs had to be honored immediately. [1] In our reading, an institution using both exempt and consent-based messages had to apply two sets of terms to the same customer record.
Systems that cannot hear a revocation
The order added a caution in a footnote. It said “callers may not deliberately design systems or operations in ways that make it difficult or impossible to effectuate revocations.” [1] The consumer lender had pointed to the mortgage servicing rule and the Fair Credit Reporting Act, which let institutions designate where certain requests and disputes are sent. [1]
The Commission was not persuaded. It said that where Congress intends to specify the means of opting out, it does so, and the statute contained no equivalent language. [1] In our reading, that put the design of the institution’s channels inside the compliance question. An interactive voice system that offered no way to say stop, or a text short code that ignored replies, could itself become evidence against the caller.
The question the ruling left open
The order addressed callers who tried to impose revocation methods on their own. [1] It did not say whether a customer and an institution could agree on a revocation method in a contract. The appeals court later noted that the Commission conceded this, and said the order did not speak to revocation procedures mutually adopted by contracting parties. [2] That gap later mattered to account agreements and patient forms.
Reassigned numbers: one call, then the caller’s risk
Whose consent counted
Several petitioners asked what happens when a consenting customer gives up a number and a carrier reassigns it. [1] One petitioner argued there is no public wireless directory and people change numbers without telling callers, so calls to reassigned numbers were inevitable. [1]
The Commission interpreted “called party” to mean the subscriber, or a non-subscriber customary user of the phone. [1] It declined to read the term as “intended recipient.” [1] It said businesses should institute new or better safeguards to avoid calling reassigned numbers. [1]
The one-call allowance
The order then gave callers a limited allowance. A caller without knowledge of a reassignment, and with a reasonable basis to believe it had consent, could make one additional call after reassignment. [1] The call served as a chance to gain actual or constructive knowledge of the change. [1] If it did not reveal the reassignment, the caller was deemed to have constructive knowledge anyway. [1]
The order said the allowance ran over an unlimited period of time. [1] It rejected the argument that the one call must connect to a person, an answering machine or voicemail. [1] It said affiliates and subsidiaries count as a single caller, so two affiliated entities could not make one call each. [1]
The burden sat with the caller. The order said the caller, not the called party, must show a reasonable basis for believing it had consent and a lack of actual or constructive knowledge of the reassignment. [1]
Seven ways to learn
The order listed methods the record showed callers could use to discover reassignments. [1]
- Include an interactive opt-out in every prerecorded or artificial-voice call, so recipients can report a wrong number.
- Record wrong-number reports received by representatives placing outbound calls.
- Let agents record new numbers when customers call in.
- Periodically ask customers by email or mail to update contact information.
- Use a dialer’s or live caller’s ability to recognize the “triple-tone” that identifies disconnected numbers.
- Set policies for deciding whether a number has been reassigned after no response to two-way calls.
- Let customers update contact information by replying to any text.
It also noted that at least one database could help determine whether a number had been reassigned. [1] It said nothing prevents a caller from requiring, by contract, that a consumer notify it when giving up a number. [1]
The Commission said the list was not exhaustive and did not imply that a reasonable caller would use any particular number of items on it. [1] It said these options, “when used as a normal practice of business, make it possible in most circumstances to comply.” [1]
What the allowance asked of an institution
On our reading, the allowance turned the institution’s customer data into the main control. A dialer that treated a number as valid until a customer complained would exhaust the one call without noticing. A system that captured wrong-number reports, disconnect tones and returned mail, and acted on them before the next campaign, stood a chance.
The affiliate rule widened the problem. A bank holding company with a card issuer, a mortgage servicer and a collections arm got one call in total, not one each. [1] In our reading, that put a group-wide number record at the center of compliance, even where business lines kept separate customer files.
Health systems faced the same issue across hospitals, clinics and affiliated physician practices. Whether affiliated entities count as one caller turned on corporate relationships the order described as affiliates and subsidiaries. [1] Organizations should ask counsel how that test applies to their own structures.
A hypothetical regional bank
Consider, as an illustration, a regional bank with a card portfolio and a collections unit. A customer gives a mobile number when opening an account, then drops it a year later without telling the bank. The carrier reassigns the number.
The card unit sends an automated payment reminder to the number. Under the 2015 order, that reminder used the group’s one allowance, whether or not anyone answered. [1] The collections unit dials the same number a week after the reminder. Unless the bank had learned of the reassignment, that second call was the bank’s risk. [1]
The illustration shows where the control had to sit. The card unit’s reminder platform needed to pass a wrong-number reply or a disconnect signal to a shared record. The collections dialer needed to read that record before dialing. In our reading, the order made those two systems one compliance system, whatever the bank’s organization chart said.
The reasoning on cost
The order set the allowance against the privacy of the new subscriber. It reasoned that the statute places no obligation on a new subscriber to answer calls or tell callers about a reassignment. [1] It noted evidence that callers sometimes continued calling after being told of a wrong number. [1]
The Commission also explained why it stopped short of strict liability. According to the appeals court, the order said the Commission could have imposed a “zero call” approach, and declined to require a result that severe. [2] That explanation became central when the allowance was challenged.
The equipment question: which dialers the rules reached
Capacity, not current use
The statute defines an autodialer by capacity. The equipment must be able to store or produce numbers to be called using a random or sequential number generator, and to dial them. [1] Three petitioners asked whether that meant present ability at the time of the call. [1]
The Commission reaffirmed its earlier view. Dialing equipment generally has the requisite capacity even if not presently used for that purpose, including when calling a set list of consumers. [1] It said callers cannot avoid the consent requirement by dividing ownership of pieces of equipment that work together. [1]
Why this mattered to institutions
A bank or health system calling its own customers or patients works from a list. A predictive dialer used for collections, or a platform that sends appointment texts, does not generate random numbers. The order’s reading meant that kind of equipment could still be an autodialer. [1]
We read that as pushing institutions toward one of two paths. Either they treated their automated platforms as regulated and built the consent records the rules required, or they routed calls through manual processes the definition did not reach. The second path cost staff time and speed. The first cost data work.
Text platforms
The order confirmed that text messages are calls under the statute. [1] It also said Internet-to-phone texts require consent. [1] An institution that moved reminders or alerts from voice to text in search of a lighter regime found the same rules applied.
Equipment split across vendors
The order closed one structural route. It said callers cannot avoid obtaining consent by dividing ownership of pieces of dialing equipment that work in concert among multiple entities. [1] An institution that licensed a dialing engine from one vendor, hosted numbers with a second and ran campaigns through a third could not, on that reasoning, treat the arrangement as outside the statute.
We read that as a procurement point as much as a legal one. The contracts for those services decided who held the call logs, the consent flags and the configuration that controlled dialing. If the institution was the caller, it needed those records whichever vendor held them.
Equipment and the exemptions
In our reading, the equipment ruling and the exemptions worked together. The broader the definition of an autodialer, the more calls needed consent, and the more the free-message exemptions were worth to the institutions that could use them.
How the order reached banks, health systems and agencies
Large banks and card issuers
The fraud and breach exemptions answered a request from the banking trade associations. [1] An issuer that already sent fraud alerts by text could keep doing so without consent, within the seven conditions. [1] Exempted messages could go only to the wireless number provided by the customer, so a number obtained from another source did not qualify. [1] In 2026 the Commission agreed with commenters that this condition unduly restricted institutions’ ability to send important, time-sensitive information, and widened it. [5]
Fraud and security teams
The order described how institutions identify events that trigger fraud alerts. It recounted the banking petitioner’s account of algorithms that flag purchases unusual for a customer, purchases of goods easily converted to cash, or a change of address followed closely by a request for new cards. [1] In our reading, the exemption made the fraud team a sender of regulated messages. Its alert logic decided which events produced exempted calls, and its tools had to respect the three-message limit per event.
The breach categories reached security and privacy teams too. Calls about a breach at a merchant, and about steps a customer could take afterward, were exempt only without marketing content. [1] A breach response plan that paired a notice with an offer of a paid monitoring product would, in our reading, take the call outside the exemption.
Credit unions and community banks
Where a smaller institution relies on a core processor or card-alert vendor for messaging, the conditions still apply to each exempted message made by the institution. [1] In our reading, the institution needed contract terms giving it the evidence the conditions required: content approvals, counts per event and opt-out logs.
Mortgage servicers and collectors
Collection calls were outside the exemptions. [1] Servicers and collectors relying on consent faced the revocation and reassignment rulings in full. [1] The affiliate rule meant one allowance per corporate family. [1]
Health systems
Hospitals gained a clear consent rule for HIPAA-subject calls and an exemption for treatment messages. [1] Billing and collection messages stayed outside the exemption. [1] In our reading, the revenue cycle and the clinical reminder program needed separate templates, separate frequency counters and separate opt-out handling.
Ambulatory networks and pharmacies
The healthcare exemption applied per provider: one message per day, up to three per week from a specific healthcare provider. [1] A network with shared scheduling across practices had to decide what counted as a specific provider for its counters. That is a question for counsel, and the answer shaped how the network built its messaging.
Public agencies
Public agencies were not the subject of the order. It noted that federal law and the Commission’s rules protect public safety answering points from robocalls that tie up first-responder resources. [1] Whether and how the statute reaches a given government caller, or a contractor calling for one, is a question for counsel.
Messaging vendors
The order addressed app providers and collect-call services as possible makers of calls. [1] It said providers that play a minimal role in sending texts are not per se liable. [1] In our reading, that left the institution that chose the content and recipients as the likely caller, and the vendor as the holder of much of the evidence.
Objections, and what the record shows
“The exemptions gave banks a free hand”
The record says otherwise. The Commission rejected the request for unlimited messages and set a three-message limit per event. [1] It barred marketing and debt collection content. [1] It required immediate opt-out. [1] The exemption was narrower than the request.
“Revocation by any means was unworkable”
The Commission said callers would not find it overly burdensome, as a general matter, to record and act on revocation requests. [1] The appeals court upheld the approach. [2] The Commission reversed course in 2026 and allowed callers to designate exclusive methods, saying that approach kept a simple, automated, easily understood method available to consumers while streamlining callers’ processing and reducing compliance burdens. [5] In our reading, both positions rest on the same premise: revocation is a data flow, and the debate was over how many inputs it must have.
“One call was arbitrary”
The appeals court held that the one-call allowance, as defended in the order, was arbitrary and capricious. [2] It set aside the treatment of reassigned numbers as a whole. [2] The Commission then built a database and a safe harbor tied to it. [3]
“The rules punished legitimate businesses”
The order addressed this argument directly. It said the statute makes it unlawful for any business, legitimate or not, to make robocalls that do not comply. [1] It said it had never viewed legitimate businesses as exempt. [1]
“The healthcare exemption should have covered billing”
The healthcare petitioner asked for account and payment calls to be included, and the Commission refused. [1] A pharmacy chain later argued in court that the partial exemption conflicted with HIPAA. [2] The court rejected that argument. It said nothing in HIPAA commands that providers be free to make autodialed calls about outstanding charges without TCPA liability. [2]
The court also found the Commission had adequately explained the line it drew between treatment messages and financial communications. [2] In our reading, that left health systems with a settled split. Clinical reminders could use the exemption; billing messages needed consent.
“Records cannot prove a negative”
The order put on the caller the burden of showing that it lacked actual or constructive knowledge of a reassignment. [1] In our reading, that was a hard burden to carry. The 2018 safe harbor addressed it by tying protection to a database check, with the burden still on the caller. [3] In our reading, proof of a negative became proof of a query.
Checks a compliance or contact-center leader could run
These checks are our suggestions, not requirements of the 2015 order or the current rules. Organizations should confirm their obligations with counsel, especially while the 2026 changes take effect.
- Consent source: for each number, record who provided it, when, through which channel and for what purpose.
- Revocation intake: list every channel where a customer or patient can say stop, and trace how each reaches every sending system.
- Category opt-outs: where opt-outs are limited to a category, store and apply them by category and account.
- Message templates: keep approved templates for exempted messages, with content checks for marketing, billing or collection language.
- Counters: count exempted messages by event and account for banks, and by day and week per provider for healthcare.
- Number hygiene: capture wrong-number reports, disconnect tones and returned mail, and act on them before the next campaign.
- Database checks: where the reassigned numbers database is used, keep the query, the date and the result for each call.
- Vendor evidence: make sure messaging contracts give the institution logs it would need to show it met each condition.
The 2015 order restated that proof of consent sits with the caller. [1] The checks above follow that line: know where each number came from, know when consent ended, and keep the record that shows both.
With hindsight
The 2015 order was challenged in the D.C. Circuit, and the court decided the case on March 16, 2018. [2] It upheld the Commission’s approach to revocation of consent through any reasonable means. [2] It also sustained the scope of the exemption for time-sensitive healthcare calls, rejecting a pharmacy chain’s challenges to it. [2]
The court set aside two other parts. It found that the Commission’s understanding of autodialer capacity would appear to bring ordinary calls from any conventional smartphone within the statute, an unreasonably expansive interpretation. [2] It found the one-call safe harbor for reassigned numbers, at least as defended in the order, arbitrary and capricious. [2] Because it could not be sure the Commission would have adopted strict liability without the safe harbor, it set aside the treatment of reassigned numbers as a whole. [2]
The court noted that the Commission was already considering a database of reassigned numbers. [2] The Commission adopted one in an order adopted December 12, 2018 and released December 13, 2018. [3] It said about 35 million numbers are disconnected and made available for reassignment each year. [3]
The 2018 order set a minimum aging period of 45 days for all numbers, a period a disconnected number must wait before reassignment. [3] It created a safe harbor for callers who check the most recent update of the database and receive a “No.” [3] The caller bears the burden of proving it checked. [3] The Commission declined to extend the safe harbor to commercial databases. [3]
The database opened later. A bureau notice published November 5, 2021 said the database was available for full use on November 1, 2021, after a beta test. [4] It set six subscription tiers. [4]
The revocation ruling changed in 2026. By then the any-reasonable-means standard sat in the Commission’s rules, at section 64.1200(a)(10). [5] In an order adopted September 30, 2026 and released October 1, 2026, the Commission let callers designate an exclusive means to revoke consent. [5] A caller may choose an automated voice or key-press mechanism, standardized words in reply to a text, or a website or phone number, if it clearly and conspicuously discloses the method. [5] Callers that do not designate one must still process requests made in any reasonable manner. [5]
The same order let callers treat a revocation as applying only to the category of informational robocalls it was directed at. [5] It said a rule treating one revocation as applying to all robocalls could stop fraud alerts, multi-factor authentication or appointment reminders the consumer still wanted. [5] Revocations of telemarketing calls still apply to all telemarketing from that caller. [5]
It also amended the 2015 bank exemption. Exempted messages may now go to numbers from a reliable source: a spouse or family member authorized on the account, a customer’s call to the institution, or another financial institution’s records. [5] The Commission said it was granting the banking association’s 2015 petition for reconsideration on that condition. [5] It encouraged institutions to use the reassigned numbers database. [5]
As checked on November 16, 2026, the rule published October 26, 2026 states an effective date of November 25, 2026. [5] The Commission sought further comment on whether to require a one-step way to revoke consent to all robocalls. [5]
What held up from 2015 is the burden. The caller still proves consent, and under the safe harbor it proves its database check. [1] [3] The healthcare exemption’s scope survived review. [2] What did not hold up is the one-call allowance and the broad equipment reading. [2] The revocation rule moved from any reasonable means toward methods a caller can automate. [5]
In our view, the lasting point is that the order made outbound contact a records discipline. Each later change altered which records a caller keeps, not whether it keeps them.
Spotted an error? Tell us and we will check it and correct the article where needed.