Net neutrality’s 2018 repeal reached the public’s end of the connection, not the agency’s circuit

When the FCC’s repeal took effect on June 11, 2018, agencies’ enterprise circuits had been outside the 2015 conduct rules. The change fell on the mass-market service residents, patients and field staff used to reach them, leaving disclosures and contracts as the controls.

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

When the FCC’s repeal took effect on June 11, 2018, agencies’ enterprise circuits had been outside the 2015 conduct rules. The change fell on the mass-market service residents, patients and field staff used to reach them, leaving disclosures and contracts as the controls.

  1. The week On Monday, June 11, 2018, the Federal Communications Commission’s Restoring Internet Freedom Order took effect.
  2. The institution’s own circuits: outside the rules before and after The first part concerns what did not change.
  3. The public’s connection: where the conduct rules had applied The second part concerns what did change.
  4. The disclosures: the one federal obligation left in place The third part concerns the transparency rule.
  5. The jurisdiction: preemption and the states as regulators and buyers The fourth part concerns which government could set rules for the public’s connection.
  6. Objections, and where the argument is weakest Four objections test the argument that the repeal mattered mainly at the public’s end of the connection.
  7. Checks for a CIO who depends on connections they do not buy The four parts translate into checks for an agency CIO, a health system’s chief technology officer or a bank’s head of digital channels.
  8. With hindsight Legal challenges to the order were consolidated in the D.C. Circuit.

The week

On Monday, June 11, 2018, the Federal Communications Commission’s Restoring Internet Freedom Order took effect. The Commission had adopted it on December 14, 2017, released it on January 4, 2018, and published it in the Federal Register on February 22, 2018. Much of it waited on one step. Its revised transparency rule contained an information collection that needed approval from the Office of Management and Budget. On May 11, 2018, the Commission announced that OMB had approved the collection for three years and that the order and its rule amendments would take effect on June 11.

The order did three things. It restored the classification of broadband internet access service as an information service, rather than a telecommunications service under Title II of the Communications Act. It reinstated the private mobile service classification of mobile broadband. And it eliminated the conduct rules of the 2015 Title II Order. In the order’s words, it found that transparency was sufficient to protect the openness of the internet and that conduct rules had greater costs than benefits.

The conduct rules had barred blocking, throttling and paid prioritization, and had included a general conduct standard. The 2018 order said the no-blocking and no-throttling rules were unnecessary. It relied instead on the transparency rule, the Commission’s enforcement authority, enforcement of ISP commitments by the Federal Trade Commission, antitrust law, consumer expectations and ISP incentives. It also said that reclassification returned jurisdiction over broadband privacy and data security to the FTC.

The transparency rule that remained, codified at 47 CFR 8.1, required providers to disclose accurate information about their network management practices, performance characteristics and commercial terms. Disclosure could be on a publicly available website or by transmittal to the Commission. The order required disclosure of congestion management, application-specific behavior, device attachment rules and security practices. It added an express duty to disclose any blocking, throttling, affiliated prioritization or paid prioritization.

The order also eliminated additional reporting obligations the 2015 order had added, returning to the disclosure requirements of the Commission’s 2010 Open Internet Order. It singled out the performance metric, which had required disclosure of packet loss and performance at peak usage times, as unduly burdensome.

The order defined its subject narrowly. Broadband internet access service was a mass-market retail service providing the capability to reach all or substantially all internet endpoints. Mass market meant services sold on a standardized basis to residential customers, small businesses and other end users such as schools and libraries. It included service bought with support from the E-rate and Rural Healthcare programs. It excluded enterprise service offerings and special access services, which the order said were offered to larger organizations through customized or individually negotiated arrangements. Public safety services as defined in the Communications Act were also outside the definition.

Finally, the order preempted state or local measures that would impose rules it had repealed or declined to impose, or more stringent requirements on any aspect of broadband service it addressed. It said this included state laws requiring disclosures of performance, commercial terms or network management inconsistent with its own.

The repeal that took effect on June 11, 2018, was argued over as if it governed the whole internet. For public agencies, hospitals and banks, it governed one part: the mass-market connection that residents, patients, customers and sometimes employees used to reach them. The circuits those institutions bought for themselves were enterprise services. Neither the 2015 conduct rules nor their 2018 repeal had applied to those.

Our argument is that the repeal moved the protection of public-facing traffic from rules to disclosures and contracts, and that the move mattered most at the public’s end of the connection. An agency that wanted to know what had changed for its own operations had to sort its services into two groups. Some depend on its own circuits, governed by contracts it negotiated. Others depend on whatever plan its residents, patients and field staff happen to buy.

The analysis has four parts. The first is the institution’s own circuits: enterprise and dedicated services, which sat outside the rules before and after. The second is the public’s connection: the mass-market service that carried residents, patients and some first responders, which is where the conduct rules had applied. The third is the disclosures: the transparency rule, the one federal open internet obligation left in place, and what an institution could do with it. The fourth is the jurisdiction: the preemption claim over state rules, and what it meant for state and local agencies as regulators and as buyers. A section on objections and a closing section of checks follow.

The parts do not overlap. The first concerns services an institution buys under its own contracts. The second concerns services the public buys. The third concerns the information providers must publish about the second. The fourth concerns which level of government could set rules for the second. Together they cover where the change landed and where it did not.

The sector here is government, because public agencies sit on both sides of the question: they are large buyers of connectivity, and they depend on residents’ connections to deliver services. Health systems and banks face the same split between their own circuits and their customers’ connections, and the checks at the end apply to them. Nothing here is legal advice about any agency’s obligations or rights.

The institution’s own circuits: outside the rules before and after

The first part concerns what did not change. The 2018 order’s definition of broadband internet access service excluded enterprise service offerings and special access services. Those are the services larger organizations buy under negotiated arrangements. A state data center’s internet transit, a hospital’s wide area network and a county’s dedicated fiber to its 911 center fall in that category if they were bought that way.

What governed those services

Enterprise services were governed by the contract between the institution and the provider. The contract set bandwidth, latency, availability, restoration times, priority and remedies. If a provider throttled an enterprise circuit, the remedy lay in the service level agreement, not in a federal conduct rule. That was true under the 2015 rules and remained true after June 11, 2018.

The FCC said as much later, in its 2020 response to the court that reviewed the order. It wrote that enterprise and non-BIAS services were not subject to the 2015 order’s restrictions. It also said that public safety entities generally purchase enterprise service contracts, citing one provider’s description of them as similar to other large agreements that government entities use to buy goods and services. The same provider said some states use master agreements negotiated through nationwide purchasing organizations.

The incident that proved the point

In the summer of 2018, a carrier slowed the mobile data service of a California county fire department while it was fighting a wildfire. Challengers raised the incident in court, and the D.C. Circuit declined to consider it because it postdated the order. The FCC’s 2020 order addressed it directly. It said the fire department’s plan from its carrier was an enterprise plan, not a mass-market service, so it was not broadband internet access service under either the 2015 order or the 2018 order.

The FCC added that even a mass-market plan might not have violated the 2015 rules, because those rules did not prohibit data caps of the kind at issue. Whatever one makes of that analysis, the factual point stands. The plan that failed the fire department was governed by its contract. The 2015 rules, had they still been in force, would not have reached it.

For an agency, the lesson of the incident is about the contract. A plan with a data cap or a speed reduction after a threshold is a plan with a data cap, whatever the regulatory regime. Emergency services need contract terms that remove or suspend those limits, in writing, before the emergency.

What the contract should say

If the contract is the control, its terms carry the weight a rule would otherwise carry. For an enterprise circuit that supports emergency or clinical services, the useful terms are specific. They include committed bandwidth, and targets for latency, jitter and packet loss suited to voice and video. They include the points where performance is measured, and remedies beyond service credits, such as termination rights for repeated failure.

Restoration needs its own terms. The FCC’s 2020 order described the Telecommunications Service Priority system, which authorizes the assignment and approval of priorities for provisioning and restoration of common-carrier telecommunication services. An agency that relies on a circuit for emergency work should confirm whether it is enrolled, and whether the contract requires the provider to honor priority in practice.

Data limits deserve explicit treatment. The wildfire incident turned on a plan whose speed was reduced after a usage threshold. A contract for emergency services can state that no usage-based reduction applies, or that limits are suspended on notice during a declared emergency. Writing that term costs little before an emergency and is hard to obtain during one.

Measuring the circuits an institution buys

A service level agreement is only as good as the measurement behind it. If the provider measures its own performance, the institution learns about failures from the provider’s report. If the institution measures at its own endpoints, it learns when its applications suffer.

Independent measurement need not be elaborate. Synthetic tests between key sites, monitoring of voice and video quality, and alerting on latency and loss give an institution its own record. When a dispute arises, that record decides whether service credits, escalation or termination rights apply. Without it, the contract’s protections depend on the provider’s account of events.

How the split differed by institution

A federal agency buying connectivity through large government contracts would be, in large part, on the enterprise side for its own operations. Its exposure to the repeal lay in the public’s access to its services. A county 911 center was in a similar position for its core network, but its field units might use commercial mobile plans, which could be enterprise or mass-market depending on how they were bought.

A rural clinic or a small library was in a different position. By the order’s definition, service bought with Rural Healthcare or E-rate support was mass-market service. For those institutions, the repeal reached their own connections as well as their patrons’. Their protection after June 2018 was the provider’s disclosure and whatever terms the program and the provider’s standard agreement offered.

A hospital system or a bank is closer to the federal agency. Its data centers and branches can run on enterprise circuits. Its patients and customers reach it from home. The split that matters is the same.

Public safety services and device connectivity

Two other categories sat outside the order’s definition, and both matter to public institutions. The order said public safety services as defined in Section 337 of the Communications Act would not meet the definition, because they are not made commercially available to the public as a mass-market service. Dedicated public safety networks therefore fell outside the debate entirely.

The order also repeated a 2010 finding that broadband internet access service does not include connectivity to one or a small number of endpoints for a particular device. Its examples included connectivity bundled with e-readers, heart monitors or energy consumption sensors. On that finding, a remote patient monitoring device with its own bundled connection fell outside the definition, to the extent the service related to the device’s functionality. Its performance would then depend on the device maker’s contract with the carrier.

What an agency should have reviewed in June 2018

The repeal gave agencies a reason to re-read their connectivity contracts, even though the contracts were not directly affected. The review was practical. It meant separating services on enterprise circuits with service level agreements from services on consumer or small business plans bought for convenience. It meant finding any contract that let the provider manage traffic in ways that could slow critical applications. And it meant checking that priority and restoration terms were written down.

Institutions that bought standardized plans were in a different position. A rural health clinic buying service with Rural Healthcare program support was, by the order’s own definition, a mass-market customer. So was a library or school buying with E-rate support. For those institutions, the repeal did reach their own connections, which leads to the second part.

The public’s connection: where the conduct rules had applied

The second part concerns what did change. The conduct rules had applied to mass-market service. After June 11, 2018, the same service was governed by the transparency rule, FTC enforcement of providers’ public commitments, antitrust law and providers’ incentives. That shift fell on the connections the public used to reach public services.

Who was on the mass-market side

The order’s definition named residential customers, small businesses, schools and libraries. It included broadband bought with E-rate and Rural Healthcare support, and service on networks supported by the Connect America Fund. For a government audience, that list covers residents reaching benefits portals, students reaching schools, patients reaching telehealth from home and small public institutions such as libraries.

The FCC’s 2020 order recognized that this side carried public safety traffic. It said the record suggested that most data communications between public safety entities and individuals likely take place over broadband internet access services, not enterprise or dedicated services. It noted that members of the public rely on broadband during emergencies to find information, and that officials gather situational awareness from residential broadband users.

Telehealth and benefits on the public side

The same order observed, in the context of the pandemic, that Americans were relying on telemedicine over mass-market broadband for routine care, triage and basic health advice. That remark was made in 2020, but the structure it describes was in place on June 11, 2018. A health system could control its own data center and its telehealth platform. It could not control the patient’s home connection, which was governed after June 2018 by the provider’s disclosures and the FTC.

The same holds for a state benefits agency, an unemployment insurance program or a motor vehicle department moving services online. Each depends on residents’ connections that the agency does not buy. Under the 2015 rules, those connections carried a federal ban on blocking and throttling. After June 11, 2018, they carried a federal duty to disclose such practices, and the consequences of breaking a published commitment.

Designing services for a connection the institution does not control

If the public’s connection is governed by disclosures rather than prohibitions, the institution’s remaining lever is the design of its own services. A service that degrades gracefully on a slow or managed connection is less exposed to any provider’s traffic management, whatever the regulatory regime.

Several design choices follow, as we see it. Forms and uploads can be resumable, so a dropped connection does not lose a benefits claim. Pages can load core functions first and heavy media later. Telehealth platforms can fall back from video to audio when bandwidth drops, with clinical rules about when an audio visit is acceptable. Alerts can go out over more than one channel, such as text messages and voice calls alongside app notifications.

These choices also help residents with weak coverage, older devices or limited data plans, which is reason enough to make them. The repeal added a regulatory reason. Where an institution cannot know how each provider will treat its traffic, it can make its traffic less sensitive to that treatment.

Measuring what the public experiences

An institution cannot manage what it does not measure. Real-user monitoring of a public portal or telehealth platform can record load times, failure rates and session drops. Segmented by the network a user connects from, those measurements show whether one provider’s customers fare worse than another’s.

That evidence has two uses. Operationally, it lets the institution tune its services for the connections its public uses. In a regulatory context, it is the kind of record that could support a complaint to the FCC or the FTC. It could also support a submission to a state attorney general if a provider’s practice harmed access to public services. Without measurement, an agency has only anecdotes.

Banks and health systems on the public side

The same split runs through financial services and health care. A bank’s branches, data centers and payment links run on enterprise circuits. Its customers reach mobile banking from home broadband and mobile plans, which are mass-market services. A health system’s hospitals run on enterprise circuits, while its patient portal, video visits and remote monitoring reach patients over whatever they buy.

For a bank, the public side carries authentication traffic as well as transactions. One-time codes by text message, push notifications and app sessions all depend on the customer’s connection. For a health system, it carries clinical traffic: a video visit, a portal message about results, or readings from a home device that uses the patient’s Wi-Fi rather than bundled connectivity. In each case the institution’s architecture choices, not its contracts, are what it controls. A bank that offers a voice call when a one-time code fails has reduced its dependence on any single network’s behavior. So has a health system that can switch a video visit to audio.

Field staff on consumer plans

There is a third group on the mass-market side: an agency’s own staff when they work on consumer plans. Caseworkers, inspectors and home health nurses may connect from home broadband or personal mobile plans. If an agency relies on those connections for field work, part of its own operations runs on mass-market service.

The FCC’s 2020 order said that public safety entities make use of dedicated or prioritized enterprise services for communications among officials, calling FirstNet perhaps the most important example of a dedicated network. It also acknowledged that public safety agencies rely on retail broadband for applications such as accessing databases and sharing data with responders. An agency that wanted its field work protected by contract had to move it onto services bought under contract.

The disclosures: the one federal obligation left in place

The third part concerns the transparency rule. After June 11, 2018, it was the remaining federal open internet rule on mass-market broadband. The order made it the foundation of its approach, and it is the part an institution could use directly.

What providers had to publish

Under 47 CFR 8.1 as revised, providers had to disclose network management practices, performance characteristics and commercial terms, sufficient for consumers to make informed choices. The order required disclosure of congestion management, application-specific behavior, device attachment rules and security practices. It also required disclosure of any blocking, throttling, affiliated prioritization or paid prioritization.

The security practices disclosure deserves attention from security teams. It tells an institution what a provider says about how it protects its network and its customers. If a provider blocks certain ports or protocols for security reasons, that practice can affect an institution’s applications as surely as a commercial practice would. The disclosure is the place to find it.

Those disclosures became the reference point for enforcement. The FCC’s 2020 order said providers know that their public statements about network management are subject to FTC enforcement, so that those statements in effect create requirements to which they are bound. It cited a memorandum of understanding under which the FTC would investigate and take enforcement action as appropriate against providers for unfair, deceptive or otherwise unlawful acts or practices.

What an institution could do with them

A disclosure is useful only if someone reads it. An agency with services that depend on residents’ connections could read the disclosures of the major providers in its area. It could note any practice that could affect its applications: video for telehealth, uploads for document-heavy benefit claims, or real-time messaging for alerts.

The disclosures also served procurement. Where an agency or clinic bought mass-market service, the provider’s disclosure described what it was buying. A procurement officer could attach the disclosure to the purchase record, so that a later change in practice was visible. Where the disclosure was silent on a point that mattered, the agency could ask for written commitments, or move to an enterprise contract.

Turning disclosures into a register

The disclosures can be organized. An institution can keep a short register of the providers that serve the bulk of its public. For each, it records the date of the current disclosure and any practice in it that bears on the institution’s services. The register needs an owner and a review date.

The register also records what the disclosures do not say. If no provider discloses how it treats real-time video during congestion, that gap is itself useful. It tells a telehealth program to plan for variable quality, and tells a procurement officer what to ask for in any contracted service.

What the FTC route required

The order placed enforcement of providers’ commitments with the FTC. The FTC acts against unfair or deceptive practices, which means its reach depends on what a provider promised and whether it kept the promise. A provider that disclosed a practice accurately was not deceiving anyone by following it.

For an institution, that shapes how to escalate a problem. A complaint is strongest when it can show a gap between a provider’s published disclosure and its conduct, supported by measurements. An institution that has read the disclosure and measured performance can make that case. One that has done neither can describe a problem but not a breach of a commitment.

The limits of disclosure

Disclosure has limits that the order acknowledged by relying on other tools. A disclosure describes practices; it does not prohibit them. A resident who cannot reach a public service because of a disclosed practice has been informed, not protected. FTC enforcement depends on a mismatch between what a provider said and what it did.

For an institution, that means disclosures support monitoring rather than guarantees. The useful question is whether the institution would notice a change. If a provider changed its practices and updated its disclosure, would anyone at a health system or benefits agency read the update? The repeal made that job worth assigning to a named person.

The jurisdiction: preemption and the states as regulators and buyers

The fourth part concerns which government could set rules for the public’s connection. The order claimed broad preemption. State and local agencies sat on both sides of that claim.

What the order claimed

The order preempted state or local measures that would impose rules the Commission had repealed or declined to impose. It also preempted measures imposing more stringent requirements for any aspect of broadband service it addressed. It said this included state disclosure requirements inconsistent with its own. The aim, in the order’s words, was to avoid a patchwork of separate and potentially conflicting requirements across jurisdictions.

For state legislatures and attorneys general, the claim set up a test. A state that wanted to restore conduct rules for its residents would have to legislate in the face of an express preemption directive. California did so in 2018. Its law and the litigation over it are part of the hindsight below.

States as buyers

States are also large buyers of connectivity, and procurement is a different lever from regulation. A state can require, as a contract term, that the providers it buys from follow particular practices on the services it buys. Whether such terms reach mass-market service the provider sells to the public, and whether preemption would apply, are legal questions for counsel. Within an agency’s own contracts, the agency sets the terms.

For an agency’s technology and procurement leaders, the practical point is that their own contracts were within their control regardless of the federal regime. Priority, throttling, data caps, restoration and reporting could all be written into enterprise agreements. The preemption fight concerned the public’s connections, where the agency was not the buyer.

What state rules mean for an institution

State rules apply to providers serving residents of the state. An institution that serves the public in several states may find that its users’ connections are governed by different rules depending on where they are. A national bank’s mobile customers, a health system that crosses state lines or a federal program delivered nationwide would each face that patchwork.

For operations, the patchwork is a monitoring question. An institution need not track every state law in detail. It needs to know, for the states where its users are concentrated, whether conduct rules apply to providers there, and to treat measurements from those states accordingly. Counsel can say which state rules are in force and what they require.

Federal agencies

In our reading, federal agencies stood outside the preemption fight. They were neither regulators of broadband nor subject to state rules on their own purchases in the same way. Their position was that of large buyers and service providers to the public. For them, the jurisdiction question reduced to the same split as the rest of this article: their own contracts on one side, the public’s connections on the other.

For federal programs delivered through states, such as benefits administered by state agencies, the picture combined both. The federal program’s design sat with the federal agency. The state agency ran the service and depended on residents’ connections in its state, under whatever state rules applied there.

Local agencies and public safety

Local agencies had a narrower set of levers, and the preemption directive aimed at local measures as well as state ones. Their levers were their own contracts, their own use of FirstNet or other prioritized services, and their participation in proceedings. Santa Clara County did participate, filing comments that the D.C. Circuit later quoted on public safety.

Objections, and where the argument is weakest

Four objections test the argument that the repeal mattered mainly at the public’s end of the connection.

The objection that enterprise services share the same network

Enterprise services ride on infrastructure that also carries mass-market traffic. The FCC itself made a version of this point in 2020. It said non-mass-market offerings often use infrastructure that also carries mass-market broadband, and quoted a wireless industry association that infrastructure built mainly for mass-market consumers increases capacity for enterprise and government users too. If mass-market traffic management changed, enterprise service on shared links could feel it.

The point holds at the level of network engineering, and it strengthens the case for contract terms. An enterprise customer’s protection against effects from shared infrastructure is its service level agreement, measured at its own endpoints. The regulatory regime for mass-market service does not supply that protection either way.

The objection that nothing visibly changed

A second objection is that, after June 2018, no wave of blocking or throttling followed. The FCC’s 2020 order said as much for public safety, citing what it called a dearth of real-world examples of public safety harms from blocking or throttling mass-market service. If nothing changed in operation, the argument about where change landed is academic.

The answer is that the argument concerns which controls applied, not whether harms occurred. An institution that relies on residents’ connections should know whether those connections are governed by a prohibition, a disclosure or a contract. The FCC has since reversed course and a court has reversed it back, as the hindsight shows, so the question has not gone away.

There is also a measurement problem in the objection. An institution that did not measure its users’ experience by network could not have seen a change in treatment by one provider. Absence of reported harm is weaker evidence when no one was looking. The FCC’s finding concerned public safety records in its proceeding; it did not establish that every public service performed the same on every network afterward.

The objection that this was a consumer issue

Another objection holds that net neutrality was a consumer and edge-provider issue, and that institutions had no particular stake. On this view, a CIO had nothing to do on June 11, 2018.

The answer is that institutions deliver services to consumers. A benefits agency, a telehealth program and a mobile bank are, in our reading, edge services, reached over mass-market connections. Their stake was the same as any online service’s, with the added weight of public obligations: access to benefits, to care and to emergency information. We would expect the greatest exposure to degraded access at institutions whose users had no alternative channel.

The objection that public safety was the real story

A fourth objection is that the most important consequence of the repeal was for first responders, and that the court’s remand on public safety proves it. The D.C. Circuit did find that the FCC had failed to consider public safety, and sent that issue back.

The remand supports the split argued here rather than undermining it. The court’s concern, as the FCC later summarized it, was the ability of first responders, critical infrastructure providers and members of the public to communicate during a crisis. The FCC’s response separated communications among officials, which ran largely on enterprise and dedicated services, from communications with the public, which ran on mass-market service. That is the same line this article draws.

Checks for a CIO who depends on connections they do not buy

The four parts translate into checks for an agency CIO, a health system’s chief technology officer or a bank’s head of digital channels.

  • Own circuits: list every service that depends on connectivity, and record whether each runs on an enterprise contract, a standardized plan or residents’ own connections.
  • Contract terms: for each enterprise contract, confirm in writing the terms on priority, throttling, data caps, restoration times and emergency suspension of limits.
  • Field staff: identify field work that relies on consumer plans, and decide whether to move it to contracted or prioritized service.
  • Public side: for services residents or patients reach from home, identify the main providers in your area and read their transparency disclosures for practices that could affect your applications.
  • Monitoring: assign someone to review those disclosures at least yearly and after any change in federal or state rules.
  • Jurisdiction: ask counsel which state rules apply to providers serving your residents today, since federal and state rules have both changed since 2018.

All but one of these checks concern contracts, inventories and monitoring rather than regulation. That follows from what June 11, 2018, changed and left alone. Where an institution buys, it can write the protection it needs. Where its public buys, it can only read, monitor and plan for a connection it does not control.

Banks and health systems can read the same checks with their own terms. For a bank, the public side is mobile and online banking, including the text messages and app sessions that carry authentication. For a health system, it is the patient portal, video visits and remote monitoring that uses patients’ own connections. For both, the enterprise side is the data centers, branches, hospitals and payment or clinical links they contract for themselves.

With hindsight

Legal challenges to the order were consolidated in the D.C. Circuit. On October 1, 2019, the court upheld the order in most respects in Mozilla Corporation v. FCC. It vacated the preemption directive, finding the Commission had not shown legal authority to issue it. It remanded three issues: the order’s effect on public safety, on pole attachments and on the Lifeline program. The FCC later noted that the court had upheld its findings on reliance on the transparency rule.

The FCC answered the remand in an order adopted October 27, 2020, published January 7, 2021, and effective February 8, 2021. It found that reclassification and the end of the conduct rules were not likely to harm public safety. It said the 2018 wildfire incident involved an enterprise plan, outside both the 2015 and 2018 orders. It also removed broadband from the services the Lifeline program supports, while preserving its authority to fund broadband through the program.

With the preemption directive vacated, states legislated. California enacted the California Internet Consumer Protection and Net Neutrality Act of 2018. Industry associations sought a preliminary injunction against it. On January 28, 2022, the Ninth Circuit affirmed the denial of that injunction, holding that once the FCC classified broadband as an information service, it lacked the authority to preempt state action of this kind.

The federal rules then reversed and reversed again. On April 25, 2024, the FCC adopted an order reclassifying broadband as a telecommunications service and reinstating bans on blocking, throttling and paid prioritization. The order also adopted enhancements to the transparency rule and reinstated a general conduct standard. It was published May 22, 2024, with an effective date of July 22. The Sixth Circuit stayed it on August 1, 2024, and, as the FCC’s bureau later put it, the 2024 rules never went into effect. On January 2, 2025, in Ohio Telecom Association v. FCC, the court set the order aside. It held that broadband providers offer only an information service, deciding the question itself without deference to the agency. It also held that the FCC could not classify mobile broadband as a commercial mobile service. Its mandate issued March 20, 2025.

In an order released July 11, 2025, and published August 8, 2025, the FCC’s Wireline Competition Bureau conformed the Code of Federal Regulations to the Sixth Circuit’s decision. It restored Parts 8 and 20 of the rules to how they would read absent the 2024 changes. In the same order it conformed Part 51 to a separate court decision, and it described the rules it removed as ones that “no longer have any operative effect.”

As of that August 2025 order, the federal arrangement looked much like the one that took effect on June 11, 2018: an information service classification, a transparency rule and no federal conduct rules. In our reading, under the Ninth Circuit’s reasoning, states such as California could set their own. What held up from June 2018 was the split this article describes. Institutions’ own circuits stayed under their contracts throughout. The public’s connections moved between federal rules, disclosures and state law. The transparency rule outlasted every change of course through that order.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • net neutrality
  • FCC
  • broadband
  • public safety
  • procurement