Why the LabMD ruling pushed regulators from “reasonable security” to named controls

In June 2018 a court voided the FTC’s order against a cancer lab because it demanded reasonable security without naming any act. Regulators answered by naming controls in orders and rules, and health organizations inherited the list and the burden of proving each control.

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

In June 2018 a court voided the FTC’s order against a cancer lab because it demanded reasonable security without naming any act. Regulators answered by naming controls in orders and rules, and health organizations inherited the list and the burden of proving each control.

  1. The week On Wednesday, June 6, 2018, the U.S. Court of Appeals for the Eleventh Circuit vacated a data security order that the Federal Trade Commission had entered against LabMD, a medical laboratory that had conducted diagnostic testing for cancer.
  2. The ruling: specific orders, open standards The first part concerns what the court actually held.
  3. The orders: naming safeguards company by company The second part concerns the FTC’s response in its own orders.
  4. The rules: from orders to regulations The third part concerns rules.
  5. The evidence: what named controls ask of a health organization The fourth part concerns what changes inside an organization when controls are named.
  6. Objections, and where the argument is weakest Four objections deserve an answer.
  7. What a health CISO can do now The four parts translate into actions that do not depend on the outcome of any rulemaking.
  8. With hindsight The FTC’s response came in drafting.

The week

On Wednesday, June 6, 2018, the U.S. Court of Appeals for the Eleventh Circuit vacated a data security order that the Federal Trade Commission had entered against LabMD, a medical laboratory that had conducted diagnostic testing for cancer. The panel, in an opinion by Circuit Judge Tjoflat, held that the order was unenforceable because it did not tell the company to stop any specific act or practice.

The facts reached back to 2005. The opinion described LabMD as a now-defunct laboratory subject to data security regulations under the Health Insurance Portability and Accountability Act. Its program, the court noted, included a compliance program, training, firewalls, network monitoring, password and access controls, antivirus software and security inspections. In 2005, contrary to company policy, the peer-to-peer file-sharing application LimeWire was installed on a computer used by the company’s billing manager. That user designated a documents folder for sharing.

Between July 2007 and May 2008, the shared folder held a 1,718-page file with personal information on 9,300 consumers. The opinion listed names, dates of birth, Social Security numbers and laboratory test codes, and for some consumers insurance details. In February 2008 a data security company downloaded the file through LimeWire and later offered LabMD its remediation services. LabMD declined and removed the application. In 2009, according to the opinion, the same company arranged for the file to be delivered to the FTC.

The Commission issued an administrative complaint in August 2013. The complaint did not allege specific unfair acts. It listed measures LabMD had failed to take. Among them were a comprehensive information security program, penetration testing, limits on employee access and training. Others were authentication measures such as two-factor authentication, updates to unsupported operating systems and controls to stop employees installing unneeded applications. An administrative law judge dismissed the complaint, finding the FTC had not proved that LabMD’s practices caused or were likely to cause substantial injury.

In July 2016 the Commission reversed. It found that LabMD had failed to implement reasonable security measures and that its practices were unfair under Section 5 of the FTC Act. Its order required LabMD to establish a data security program that met the Commission’s standard of reasonableness. The order was to run until July 28, 2036, or twenty years from the latest federal court complaint alleging a violation, whichever came later. LabMD petitioned the Eleventh Circuit for review and asked it to stay the order. It argued that compliance was unfeasible given its defunct status and minimal assets. The court granted the stay while it considered the petition.

The June 6 opinion assumed, without deciding, that LabMD’s negligent failure to maintain a reasonable program was an unfair practice. It then held the remedy defective. The court wrote that the order “contains no prohibitions.” Instead, it said, the order commanded LabMD “to overhaul and replace its data-security program to meet an indeterminable standard of reasonableness.” The court explained that such an order would leave a district court, in any enforcement proceeding, to manage the overhaul itself. It granted LabMD’s petition and vacated the order.

The LabMD ruling is easy to misread as a ruling that companies need not keep reasonable security. It held nothing of the kind. The court assumed that a negligent failure to keep reasonable security could be an unfair practice. What it rejected was an order that enforced reasonableness as a standard without saying what the company must do or stop doing. The question it left regulators was a drafting question: how to name the conduct.

Our argument is that regulators answered by naming controls, and that health organizations ended up holding the list. In January 2020 the FTC said it was writing specific safeguards into its data security orders, and that it had been mindful of LabMD in doing so. In December 2021 it wrote specific safeguards into a rule for financial institutions. In January 2025 the Department of Health and Human Services proposed to do the same in the HIPAA Security Rule. It cited a court decision on encryption that, in HHS’s account, read the existing rule as asking only for a mechanism.

The analysis has four parts. The first is the ruling: what the court decided about orders and what it left open about standards. The second is the orders: how the FTC changed its data security orders afterward. The third is the rules: how specificity moved from individual orders into regulations, including the proposed HIPAA changes. The fourth is the evidence: what a named-control regime asks of a health system, an ambulatory network or a business associate. A section on objections and a closing section on what to do follow.

The parts do not overlap. The ruling concerns the legal requirement. The orders concern the FTC’s remedies against individual companies. The rules concern obligations that apply to everyone in a class. The evidence concerns what an organization must be able to show once controls are named. A health organization can be touched by all four: by HIPAA as a covered entity or business associate, and by the FTC where its activities fall within the FTC Act.

Nothing here is legal advice. Which regime applies to an organization, and how, depends on facts that counsel should assess.

The ruling: specific orders, open standards

The first part concerns what the court actually held. The opinion is narrower than its reputation, and the narrowness matters for what came after.

The assumption the court made

LabMD raised several grounds for setting aside the order. The court addressed one: whether the order was enforceable. On the merits it made an assumption. It said it would assume the Commission was correct that LabMD’s negligent failure to design and maintain a reasonable program invaded consumers’ privacy and was an unfair act or practice.

That assumption is the key to reading the case. The court did not hold that the FTC lacked authority over data security, or that reasonableness was too vague to serve as a legal standard of liability. It left both questions undecided. An organization that read LabMD as a release from reasonable security was reading something the court did not write.

Why the order failed

The court’s reasoning ran through enforcement. A cease and desist order is enforced in federal district court, through civil penalties or contempt. The court compared FTC orders with injunctions and said the prohibitions in both must be specific, or they may be unenforceable. A footnote noted that penalties for violating FTC rules had been raised to $41,484 per violation.

The LabMD order, the court said, contained no prohibitions. It required a program reasonably designed to the Commission’s satisfaction. In an enforcement proceeding, the Commission would have to show the program was unreasonable, and the court would have to decide what reasonable meant. The opinion said the order effectually charged the district court with managing the overhaul, a scheme Congress could not have envisioned.

The two routes the court described

The opinion also described the FTC’s two ways of defining unfair practices. The first is rulemaking: the Commission may prescribe rules that define unfair acts with specificity, and a rule becomes, in the court’s words, in essence an addendum to Section 5. The second is case-by-case litigation, before an administrative law judge or in district court. Either route ends in an order or rule that a court will later be asked to enforce.

Read this way, the ruling set a drafting requirement for both routes. An order had to name acts. A rule could name them for a whole class of companies. Both answers appeared later.

The FTC used each route in turn. Its 2019 orders took the case-by-case route and named safeguards company by company. Its 2021 Safeguards amendments took the rulemaking route and named elements for every financial institution under its jurisdiction. HHS, which enforces HIPAA through its own rules, proposed the rulemaking route for health care. The sequence follows the map the court drew.

Why specificity protects the regulated as well

The ruling reads as a loss for the regulator. It also describes a benefit for the regulated. An order that names acts tells a company what compliance looks like. An order that demands reasonableness leaves the company to guess, and then to defend its guess in a contempt proceeding where penalties attach to each violation.

For a security leader, that difference is concrete. Under a named order, the program can be built around a list, with an owner and evidence for each item. Under a reasonableness order, the program must anticipate what a court would later find reasonable, with no list to test against. The court’s point about the district judge applies to the company as well: nobody could tell in advance what the order required.

What the complaint already contained

The irony of the case is in the 2013 complaint. The court quoted the FTC’s list of what LabMD did not do. It did not use penetration tests to assess risks. It did not require two-factor authentication or prohibit password reuse for remote users. It ran operating systems the vendor no longer supported. It did not stop employees installing applications they did not need, or adequately review records of activity on its network.

Those are specific controls. The complaint named them as evidence of unreasonableness, and the order then fell back on reasonableness as the remedy. The specificity the court demanded was sitting in the Commission’s own pleading. The later shift, in our reading, consisted largely of moving that list from the complaint into the order.

The orders: naming safeguards company by company

The second part concerns the FTC’s response in its own orders. The response was public, and it named LabMD.

What the FTC said it changed

On January 6, 2020, the director of the FTC’s Bureau of Consumer Protection, Andrew Smith, published a post on the agency’s business blog about new data security orders. The post said strengthening the agency’s data security orders had been one of its leaders’ first priorities. It said the agency had been mindful of the Eleventh Circuit’s 2018 LabMD decision, which struck down an FTC data security order as unenforceably vague.

The post described three groups of changes. First, orders would still require a comprehensive, process-based program, but would also require specific safeguards tied to the problems alleged in each complaint. Examples included yearly employee training, access controls, monitoring for incidents, patch management and encryption. Second, third-party assessors would face more accountability: more rigorous reviews, retention of documents, and FTC approval of assessors every two years. Third, companies would present their written program to the board or governing body each year, and a senior officer would certify compliance to the FTC annually.

Why the changes answered the court

Each change addressed a weakness the court identified. Named safeguards gave an enforcing court something to test: whether the company encrypted, patched or trained, rather than whether its program was reasonable. Assessor accountability gave the FTC a record of whether those safeguards operated. Board presentation and officer certification put a named person on the record for the program’s state.

The changes did not discard the older structure. Orders still required a comprehensive information security program, built on risk assessment and adjusted as the business changed. The named safeguards sat inside that program as minimums. A company could not satisfy an order by meeting the list alone, and could not escape the list by pointing to its risk assessment.

The post said seven orders announced in 2019, against an array of diverse companies, reflected the new approach. The orders still contained process requirements, including risk assessments and a written program. The difference was that the process now sat beside a list a court could read.

Assessors and their working papers

The assessor changes deserve close attention. Under the post’s description, assessors would have to conduct more rigorous reviews, using independent sampling, employee interviews and document review. They would have to keep their assessment documents and could not refuse to give them to the FTC on certain privilege grounds. The FTC could approve and re-approve assessors every two years and require a company to replace one that performed poorly.

Those terms change the relationship between a company and the firm that assesses it. An assessment becomes a record the regulator can read, not a report the company receives. For a security team, that means the evidence it shows the assessor is, in effect, evidence shown to the FTC. Sampling and interviews test whether controls operate, rather than whether policies exist.

What an order looks like from inside a company

Taken together, the changes produce a recognizable operating model. Each named safeguard needs an owner. Each owner needs to produce evidence that the safeguard operated through the period. The assessor samples that evidence. The written program goes to the board each year. A senior officer signs a certification to the FTC, and will want sub-certifications from the owners before signing.

That model is familiar from financial reporting controls. The 2020 post drew a related comparison, likening officer certification to certification requirements in securities law. The effect is to make security evidence flow up a chain of named people, each accountable for a piece.

What it meant for health companies under FTC jurisdiction

For organizations outside HIPAA, or with activities outside it, Section 5 of the FTC Act can apply to data security. Health apps, consumer health platforms and vendors that handle health data can fall there, depending on facts counsel should check. After 2019, a settlement with the FTC on data security would come with a list of required controls, a named assessor and annual certification by a senior officer.

For HIPAA covered entities and business associates, the effect was indirect but real. The FTC’s orders showed how a regulator writes enforceable security obligations. The list in the 2020 post overlapped with areas HIPAA’s Security Rule already addressed through standards and implementation specifications, such as access control and encryption.

The rules: from orders to regulations

The third part concerns rules. The court had pointed to rulemaking as the route for defining unfair practices with specificity across a class. Two rulemakings followed that route, one final and one proposed.

The FTC Safeguards Rule

On December 9, 2021, the FTC published amendments to its Standards for Safeguarding Customer Information, which apply to financial institutions under its jurisdiction. The Commission described five main changes. The first added provisions giving covered institutions more guidance on specific aspects of a program, such as access controls, authentication and encryption. The second added accountability measures, such as periodic reports to boards of directors or governing bodies.

That rule does not apply to hospitals or health systems as such. It matters to them in two ways. Some organizations in health care finance, such as entities that extend credit to patients, may fall within its scope, a question for counsel. And it showed a regulator converting the order-by-order approach into a regulation with named controls and board reporting, the template HHS would later propose for HIPAA.

Breach reporting joins the list

The FTC added another named obligation in a rule published November 13, 2023. It required financial institutions under the Safeguards Rule to report to the Commission any notification event in which unencrypted customer information involving 500 or more consumers is acquired without authorization. The amendment took effect on May 13, 2024.

The reporting rule shows how named controls and named duties reinforce each other. The trigger turns on whether the information was encrypted. An organization that can prove encryption of a given data set has a different reporting position from one that cannot. Encryption stops being only a safeguard and becomes a fact with legal consequences, and encryption was also the issue in the M.D. Anderson case that HHS later cited.

The HIPAA Security Rule as written

HIPAA’s Security Rule took a different shape from the start. Its standards are mandatory, but some implementation specifications are labeled addressable. Under 45 CFR 164.306(d), as it reads today, a regulated entity must assess whether each addressable specification is reasonable and appropriate. If it is, the entity implements it. If not, the entity documents why and implements an equivalent alternative measure if reasonable and appropriate.

Encryption of electronic protected health information is one of those addressable specifications. Section 164.312 requires covered entities and business associates to implement a mechanism to encrypt and decrypt it. That wording became the subject of litigation over a civil money penalty against the University of Texas M.D. Anderson Cancer Center.

The court decision HHS cited

In its January 2025 proposal, HHS described the Fifth Circuit’s 2021 decision in that case. It said the court held, among other things, that the Security Rule does not say how effective an encryption mechanism must be. Under the court’s interpretation, HHS wrote, a regulated entity could meet its encryption obligations by implementing a mechanism without regard for the effectiveness of the implementation. HHS also noted the court’s view that a mechanism could include directing employees to sign an agreement requiring encryption of portable devices.

HHS said it disagreed with the court’s view that requiring employees to sign an agreement to encrypt portable devices was enough, and that additional clarity was warranted. The parallel with LabMD is direct, in our reading. In each case a court read a security obligation more narrowly than the agency did, because of how the obligation was written. In each case the agency’s response was to write the obligation more specifically.

Encryption as the test case

Encryption became the hinge for a reason. It is addressable under the current rule, so an entity may document why it is not reasonable and appropriate and adopt an alternative. And the specification asks for “a mechanism” to encrypt and decrypt, without saying where it must be deployed or how well it must work. Both features gave room for the reading HHS later objected to.

HHS’s proposal responded to both. It would end the addressable category. And in discussing the court decision, it said additional clarity was warranted so that regulated entities understand their obligation to have encryption mechanisms in place and deployed throughout the enterprise. A mechanism on paper would no longer be enough, in HHS’s framing; deployment would be the test.

For a health organization, the practical question is where encryption is missing today and why. Portable devices, removable media, backups, interfaces between systems and data held by vendors are the places to look. Each gap is either a project or a documented exception with a stated reason.

Who the proposal would reach

The proposal applies to covered entities and business associates, which HHS called regulated entities. Its impact analysis said this includes government entities that meet the definition of a covered entity, giving state Medicaid agencies as an example. A state agency running a Medicaid program would therefore face the same named controls as a hospital, if the rule were finalized as proposed.

The reach matters for public agencies that may think of HIPAA as a provider’s problem. A Medicaid agency’s eligibility and claims systems, its contractors and its data exchanges would all fall within the proposed asset inventory and network map. So would its business associates, who would owe it verification of their technical safeguards.

What the HIPAA proposal would name

The proposal, published January 6, 2025, would remove the distinction between required and addressable implementation specifications. HHS said that eliminating the distinction would make clear what had always been a requirement: that the Security Rule sets a floor. The flexibility would remain in how an entity meets the standards, not whether it meets them.

The proposal also named controls. It would require a written technology asset inventory and a network map showing how electronic protected health information moves through an entity’s systems. Its regulatory impact analysis listed what each regulated entity would need to do. The list included a compliance audit, multi-factor authentication, penetration testing, network segmentation, disabling unused ports and removing extraneous software. It also included updated policies, revised business associate agreements and new workforce training. Business associates would have to verify their technical safeguards to covered entities.

Set that list beside the 2013 LabMD complaint. Penetration testing, two-factor or multi-factor authentication and removal of unneeded software appear in both. The health sector’s own regulator proposed to name in its rule controls the FTC had once cited only as evidence of unreasonableness.

The evidence: what named controls ask of a health organization

The fourth part concerns what changes inside an organization when controls are named. Under a reasonableness regime, the central record is the risk analysis and the reasoning behind each choice. Under a named-control regime, the central record is proof, control by control, that each one operated.

From rationale to proof

The addressable structure rewarded good documentation of judgment. An entity that decided encryption was not reasonable for a system could record why and adopt an alternative. In an investigation, that documentation was the defense. Under the proposed rule, the same entity would need evidence that encryption was deployed, or that a specific exception applied with compensating controls.

The shift changes who produces the evidence. Rationale is written by compliance and security leaders. Proof of operation is produced by systems: device management reports, identity provider logs, scan results and segmentation tests. Health organizations that kept the rationale in policy documents and the proof nowhere in particular would need to connect the two.

Inventory as the foundation

Named controls are only testable against a known population. HHS said the proposed asset inventory forms the foundation for an accurate risk analysis. In practical terms, an organization cannot show that every system holding electronic protected health information uses multi-factor authentication unless it knows which systems those are.

For a health system, the inventory spans clinical applications, imaging and laboratory systems, medical devices, cloud services and the business associates that run systems on its behalf. For an ambulatory network, it spans practice management and electronic health record platforms, which may be hosted by vendors, and the devices staff use. The LabMD file sat on a billing workstation. A billing workstation belongs on the inventory as surely as an electronic health record server.

Exceptions and compensating controls

Named controls create a new category of record: the exception. The proposal’s discussion of multi-factor authentication described exceptions and required reasonable and appropriate compensating controls where one applied. Each exception becomes something an investigator can ask about. When was it granted, by whom, for which system, with what compensating control, and when will it end?

Legacy clinical systems are likely to generate exceptions. So are medical devices that cannot accept new authentication methods. A health organization that tracks exceptions in a register, with owners and end dates, can answer those questions. One that grants exceptions informally cannot.

Business associates and verification

The proposal would require business associates to verify their technical safeguards to covered entities, and covered entities to obtain that verification. That turns a contractual promise into an evidentiary exchange. A covered entity would need to request, receive, review and file verification from each business associate, and act when it is missing or weak.

The FTC’s 2019 orders had already moved in this direction for assessors. They required more rigorous independent assessments and access to assessors’ documents. HIPAA’s proposal applied a similar idea to the supply chain. The organizations doing the verifying would be hospitals and clinics, not a federal agency.

How the burden differs by organization

A large health system has a security function, a compliance office and internal audit. Its challenge under named controls is scale: thousands of systems, a long vendor list and clinical technology that changes slowly. Its evidence will come from tools that already exist, if those tools cover the whole inventory.

An ambulatory network or physician group may have a small team and rely on vendors for its core systems. Its challenge is reach. It must obtain evidence from vendors that host its electronic health record and practice management systems. Under the proposal, that evidence would arrive as business associate verification.

A business associate faces the burden from both sides. It must verify its own safeguards to each covered entity it serves, and obtain verification from its subcontractors. A billing company, a cloud host or an analytics vendor serving dozens of providers would answer the same questions repeatedly unless it prepared a standard evidence package.

What an investigator would test

Named controls change the shape of an investigation. An investigator can pick a sample of systems from the inventory and ask, for each, whether multi-factor authentication is enforced, whether encryption is deployed, and when it was last scanned. An organization either has the records or does not.

That favors organizations whose controls generate records automatically. It disfavors those that rely on attestations by system owners. The LabMD complaint itself listed what the company had not done: test its network for vulnerabilities, control installations and review activity records. Under a named-control regime, those are the first questions, not the last.

Sequencing the work

Named controls arrive as a list, but the work has an order. In our reading the inventory comes first, because every other control is tested against it. Multi-factor authentication on remote access and email comes next, because both face the internet and the control is well understood. Encryption of portable devices and backups follows, then segmentation and the removal of unused services, which take longer and touch clinical operations.

Exceptions run alongside all of it. Each time a system cannot take a control, the exception register records why and what compensates. Business associate verification can start at once, because it depends on vendors’ answers rather than internal projects. None of the steps requires a final rule to begin.

Board reporting

Both the FTC’s orders and its Safeguards Rule required reporting to a board or governing body. Board attention also figured in the 2020 FTC post, which cited research linking board-level attention to fewer breaches. A board that receives a report built on named controls can ask direct questions: which controls are deployed, where are the gaps, and which exceptions are open. A report built on reasonableness invites only a general reassurance.

The report’s format matters as much as its content. A table that lists each named control, its coverage across the inventory, the open exceptions and the date of the last test gives directors something to compare from one meeting to the next. Trends become visible: coverage rising, exceptions aging, tests slipping. A narrative report cannot carry that history, and a board cannot hold management to a trend it never sees.

Objections, and where the argument is weakest

Four objections deserve an answer.

The objection that named controls go stale

Lists of controls can age. A rule that names today’s safeguards may lag tomorrow’s threats, and the flexibility of a reasonableness standard was designed to avoid that.

The objection is fair, and it explains why both the FTC and HHS kept process requirements beside the named controls. A risk analysis still decides what else is needed. The named controls set a floor, which is how HHS described the rule’s purpose. A floor can be raised by later rulemaking. A standard that no court will enforce provides no floor at all.

The objection that lists invite box-checking

A related objection holds that named controls encourage organizations to satisfy the list and stop thinking. A hospital could deploy multi-factor authentication everywhere the rule names and still miss the exposure that matters to it.

The answer lies in how the regimes combine lists with process. The FTC’s orders kept a comprehensive program and risk assessment beside the named safeguards. HHS’s proposal kept the risk analysis and added an inventory to make it more accurate. A list without a risk analysis invites box-checking. A risk analysis without a list, as LabMD showed, may be unenforceable.

The objection that the LabMD facts were unusual

A third objection is that LabMD was a small, defunct company with an unusual history, and that general lessons should not be drawn from it. The opinion itself describes a company that had stopped operating and a data exposure discovered by a firm selling remediation services.

The facts were unusual. The holding was general. It concerned the drafting of FTC orders, and the FTC treated it as general, saying so when it changed its orders. HHS’s later reasoning about the M.D. Anderson decision followed the same pattern on different facts.

The objection that the HIPAA proposal may not become final

A fourth objection is that the HIPAA proposal remains a proposal. As checked on November 15, 2026, the Federal Register showed the January 2025 proposal and no final rule amending the Security Rule. Organizations cannot be held to proposed requirements.

Proposed requirements bind no one, and the argument here does not depend on the proposal becoming final. It depends on the direction regulators took after a court refused to enforce a general standard. The FTC orders and the Safeguards Rule apply to those they cover. The HIPAA proposal is HHS’s public statement of the controls it wants regulated entities to have, and of its view that the rule has always set a floor. An organization deciding what to do now can read that statement for what it is.

What a health CISO can do now

The four parts translate into actions that do not depend on the outcome of any rulemaking.

  • Ruling: confirm with counsel which regimes apply to each line of business, including any activities that fall under the FTC Act rather than HIPAA.
  • Orders: compare your control set with the specific safeguards the FTC has written into its orders since 2019, and note any you lack.
  • Rules: map each addressable HIPAA specification to your decision and its evidence, and mark each one where the decision was not to implement.
  • Inventory: maintain a technology asset inventory and network map covering every system that creates, receives, maintains or transmits electronic protected health information, including workstations used for billing.
  • Exceptions: keep a register of control exceptions with an owner, a compensating control and an end date for each, and review it quarterly.
  • Business associates: request written verification of technical safeguards from each business associate now, and record which ones cannot provide it.
  • Encryption: list every place electronic protected health information is stored or moves without encryption, and decide for each whether it is a project or a documented exception.

Public agencies that are covered entities, such as a state Medicaid agency, can run the same checks. Their inventories will include contractors’ systems and data exchanges with providers and plans. The verification step carries particular weight where an agency relies on contractors for core systems.

The order the court struck down in 2018 asked LabMD for reasonableness and named nothing. The documents that followed named controls. A health organization that can show, control by control, what it runs and where its exceptions are will be ready for either kind of regulator.

With hindsight

The FTC’s response came in drafting. On January 6, 2020, the director of its Bureau of Consumer Protection wrote that the agency had strengthened its data security orders, and that it had been mindful of the Eleventh Circuit’s 2018 decision. The new orders kept a comprehensive program but added specific safeguards tied to each complaint, tighter oversight of third-party assessors, annual board presentations and annual certification by a senior officer.

Rules followed. On December 9, 2021, the FTC published amendments to its Safeguards Rule for financial institutions under its jurisdiction. They added guidance on specific elements such as access controls, authentication and encryption, and required periodic reports to boards or governing bodies. The amendments took effect on January 10, 2022, and the FTC later delayed the effective date of portions of them in a final rule published November 23, 2022. A further amendment, published November 13, 2023, and effective May 13, 2024, required institutions to report to the FTC notification events involving unencrypted information of 500 or more consumers.

The 2021 Safeguards amendments also reached further. The FTC expanded the definition of financial institution to cover “finders,” companies that bring buyers and sellers together, and exempted institutions that collect less customer information from certain requirements. Named controls were paired with a scaled burden for smaller firms.

Health care followed more slowly. On January 6, 2025, HHS proposed to amend the HIPAA Security Rule. The proposal would remove the distinction between required and addressable implementation specifications, require a technology asset inventory and network map, and require controls including multi-factor authentication, penetration testing and network segmentation. HHS said recent court decisions affected enforcement of the rule. It described the Fifth Circuit’s 2021 ruling in the M.D. Anderson case on encryption and said it disagreed with the court’s view that an employee agreement to encrypt portable devices was enough. Comments were due March 7, 2025. HHS stated that the proposal, if adopted, would impose mandates resulting in expenditures of more than $183 million in any one year by state, local and tribal governments in the aggregate, or by the private sector.

The proposal’s impact analysis said it would reach government entities that meet the definition of a covered entity, such as state Medicaid agencies, as well as providers, plans and business associates. As checked on November 15, 2026, the Federal Register showed no final rule adopting the proposal. The current eCFR text of 45 CFR 164.306 still divides implementation specifications into required and addressable, and section 164.312 still lists encryption and decryption as addressable. The same section still requires procedures to verify that a person or entity seeking access is the one claimed, without naming multi-factor authentication.

What held up from June 2018 was the court’s drafting point. Regulators that wanted enforceable security obligations named controls, in orders and then in rules. What remains open, for HIPAA entities, is whether the named list becomes binding by rule. In our reading, the safer assumption for a health organization is that future enforcement will ask for evidence control by control. That is the form each regulator in this story adopted once a court asked for specificity.

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

  • LabMD
  • FTC
  • HIPAA Security Rule
  • data security orders
  • encryption