BLU’s 2018 FTC order treated a phone’s update software as a supplier someone had to oversee
The FTC’s April 2018 complaint against phone seller BLU listed four failures in overseeing the company that pushed its firmware updates. Read as a procurement checklist, the four steps fit any bank, health system or agency that buys devices.
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 FTC’s April 2018 complaint against phone seller BLU listed four failures in overseeing the company that pushed its firmware updates. Read as a procurement checklist, the four steps fit any bank, health system or agency that buys devices.
- The week On Monday, April 30, 2018, the Federal Trade Commission announced a proposed settlement with BLU Products, Inc., a Florida-based seller of mobile phones, and its co-owner and president.
- Selection: an update service chosen for a function, not for its access The complaint said BLU “failed to assess or evaluate the privacy or security practices of ADUPS prior to entering into an agreement with that company.”
- Standards: no written rules for watching the supplier The second failure in the complaint was the absence of “written data security standards, policies, procedures or practices that apply to the oversight of their service providers, including ADUPS.”
- Contract: an agreement that bought updates and said nothing about data The third failure was “failing to contractually require their service providers to adopt and implement data security standards, policies, procedures or practices.”
- Assessment: nobody evaluated the code that ran with root access The fourth failure was “failing to adequately assess the privacy and security risks of third-party software, such as ADUPS.”
- Why the update channel concentrated all four failures The four failures are separate duties, but in the BLU complaint they met in one component.
- What the case did not decide The BLU matter was a settlement.
- What the order added: consent, assessments and records The proposed order went beyond the four failures.
- How the four steps fall on banks, health systems and agencies The duties that apply to a buyer differ by sector, and organizations should confirm with counsel which apply to them.
- Objections: scale, cost and who can inspect firmware Three objections deserve an answer.
- What a procurement and security leader could check The complaint’s four failures translate into checks that need no new budget line, only an owner.
- With hindsight The FTC finalized the BLU order on September 10, 2018.
The week
On Monday, April 30, 2018, the Federal Trade Commission announced a proposed settlement with BLU Products, Inc., a Florida-based seller of mobile phones, and its co-owner and president. The Commission alleged that they had misled consumers about privacy and data security. The vote to accept the proposed consent order for public comment was 2-0, and the order was open for comment through May 30, 2018. The Commission published the complaint and a proposed decision and order the same day.
The complaint described a business built on other firms. BLU called itself a mobile manufacturer, but it outsourced manufacturing to original device manufacturers, which built BLU-branded devices to its instructions and purchase orders. According to the complaint, BLU chose certain software that came preinstalled on the devices, the default settings consumers first saw and certain security features. The complaint said that, since at least 2015, BLU had licensed software from ADUPS Technology Co., Ltd., a China-based company, to provide firmware updates, and had directed its manufacturers to preinstall it.
The complaint said the arrangement gave ADUPS “full administrative access and control” of the devices. BLU had contracted with ADUPS only for firmware over-the-air updates, the service that delivers security patches and operating system upgrades over wireless and cellular networks. Yet until at least November 2016, the complaint said, the ADUPS software sent personal information to ADUPS servers without consumers’ knowledge or consent. It listed full contents of text messages, real-time cell tower location data, call and text logs with full telephone numbers, contact lists and lists of installed and used applications. Text messages went to the servers every 72 hours and location data every 24 hours.
Reports of the collection became public on or about November 15, 2016, according to the complaint. Some consumers then stopped using their devices. Others disabled the ADUPS software and, the complaint said, were left with devices unable to receive critical security updates. BLU posted a notice saying ADUPS had updated its software to stop the collection. The complaint alleged that BLU nonetheless continued to let ADUPS operate on older devices without adequate oversight.
The Commission’s case rested on two statements in BLU’s privacy policy. One said service providers had access to the personal information needed to perform their services. It added that they “may not use it for other purposes.” The other promised “appropriate physical, electronic, and managerial security procedures.” Count I alleged the first was false, because ADUPS received data it did not need for updates. Count II alleged the second was false. To support it, the complaint listed four oversight failures. It alleged that BLU had not performed adequate due diligence in choosing and keeping service providers and had not adopted written standards for overseeing them. It also alleged that BLU had not required them by contract to protect data, and had not adequately assessed the risks of third-party software.
The complaint added that preinstalled software on BLU devices contained commonly known vulnerabilities, including susceptibility to command injection attacks. It said an unknown third party could exploit them to gain full access to a device, factory reset it, record its screen and install malicious applications.
The proposed order barred misrepresentations about privacy and security. It required a documented information security program covering the development and management of devices, with seven listed elements, one of them the selection, retention and contracting of service providers. It required independent assessments for the first 180 days and every two years for 20 years. Before collecting location data or the content of communications, the respondents would have to disclose the practice outside any privacy policy and obtain affirmative express consent. The order recorded that the respondents neither admitted nor denied the complaint’s allegations.
Enforcement against a phone seller can look like a consumer protection story with little to say to a bank, a health system or a public agency. We read the BLU complaint differently. Paragraph 21 of the complaint is a procurement checklist written in the negative. It names four things a company should do before and after it lets another company’s code run on the devices it sells. Every organization that buys, issues or permits mobile devices depends on the same kind of chain, whether or not it can see it.
The chain had three links that mattered. BLU chose the software. Contract manufacturers installed it. A third company, ADUPS, ran the update service with administrative control over BLU’s devices, according to the complaint. An organization buying BLU phones would have dealt with the first link, perhaps through a reseller. It would have had no contract with the second or third. Yet the third link held administrative control of the device.
We take the case in four parts, following the four failures the complaint listed. They divide the life of a supplier relationship without overlap:
- Selection: whether anyone assessed the supplier’s privacy and security practices before signing.
- Standards: whether written rules governed how the supplier would be overseen once signed.
- Contract: whether the agreement itself required the supplier to protect data.
- Assessment: whether anyone evaluated the risk of the supplier’s software as it actually ran.
Our argument is that the update channel was the place where all four failed together, and that, as we see it, it is still a link buyers can easily leave unchecked. After the four parts we explain why the update channel concentrated them and what the case did not decide. We then look at what the order added, how the steps fall on each sector, the objections and what a leader can check.
Selection: an update service chosen for a function, not for its access
The complaint said BLU “failed to assess or evaluate the privacy or security practices of ADUPS prior to entering into an agreement with that company.” The company had a narrow need. It wanted security patches and operating system upgrades delivered over the air. It contracted for that service and, according to the complaint, asked ADUPS to perform no other.
The difficulty is that the function and the access did not match. Delivering firmware requires the ability to change the device at its lowest level. The complaint described the result as full administrative access and control. A supplier chosen for a modest service therefore held the broadest privileges on the product. The complaint also described ADUPS as a company that offered advertising and data mining services as well as firmware updates. In our reading, that mix of business lines was the kind of fact a selection review exists to surface.
A selection review can be scoped by the service purchased, so that a vendor supplying a scheduling tool gets a scheduling-tool review. The BLU facts show why the scope should follow the access granted instead. A firmware update service is a supplier with root access, whatever the contract calls it. A review scoped to the service would ask whether patches arrive on time. A review scoped to the access would ask what else the code can do, what it sends home, where its servers sit and who else the supplier works for.
For a buyer of devices, the selection question moves up one level. A bank or agency does not choose the update service on a phone. It chooses the phone. So its selection review has to ask how the device brand chooses its own suppliers. That is a different question from asking whether the device meets a feature list. It asks whether the brand has a process at all, and whether the brand will describe it.
What a selection file would contain
A selection file for a device line can be short. It names the preinstalled software that runs with system privileges and the company behind each component. It records where each component sends data and why. It records whether the brand assessed those suppliers, and when. It records the update path: who signs updates, who pushes them and how long the brand commits to supplying them. Gathering it does not require access to source code. It requires a brand willing to answer, and a buyer willing to walk away from one that will not.
Questions a brand should be able to answer
A buyer’s questions to a device brand can be few and specific. Which preinstalled applications and services hold system-level permissions. Which of them the brand wrote and which it licensed. Which companies operate the servers those components talk to, and in which countries. Whether the brand reviewed each supplier’s privacy and security practices before signing, and whether it can say when. How the brand would learn that a component was sending data it should not.
The answers matter less for their content than for their existence. A brand that cannot list its own privileged components has not done the first step the complaint described. A buyer that accepts that answer has inherited the gap. Recording the answers also gives the buyer something to compare against later. If a public report alleges that a component collects message contents, the buyer can check whether the brand ever disclosed that component and what it said the component did.
Selection has a timing problem that the BLU facts make visible. The complaint said BLU had licensed the ADUPS software since at least 2015. A buyer that chose BLU devices in that period would have done so before the November 2016 reports. Selection diligence, like the contract, happens before anything has gone wrong. It therefore depends on asking questions whose answers seem unnecessary at the time.
Standards: no written rules for watching the supplier
The second failure in the complaint was the absence of “written data security standards, policies, procedures or practices that apply to the oversight of their service providers, including ADUPS.” This failure is separate from selection. A company can choose a supplier carefully and then stop looking at it. Written standards are what turn a one-time decision into a continuing duty with an owner.
The complaint’s account of events after November 2016 shows the cost of the gap. Once reports of the collection became public, BLU told consumers that ADUPS had updated its software to stop. The complaint alleged that BLU still let ADUPS operate on older devices “without adequate oversight.” A written standard would have had to say what oversight meant after an incident. It would name which device models were checked, by whom and against what evidence, and who signed off that the collection had stopped on each.
Written standards matter for a second reason. They are what an examiner, auditor or regulator reads. A company that says it oversees its suppliers, as BLU’s privacy policy in effect did, needs a document that shows how. Without one, the promise and the practice cannot be compared, and the gap between them is what the Commission alleged was deceptive. That was the structure of Count II.
Standards for a supplier the buyer cannot see
For an organization buying devices, the standard has to cover suppliers at one remove. The organization cannot audit ADUPS. It can write down what it requires of device brands in their own oversight of preinstalled software, and on what schedule it asks for evidence. It can also write down what it does when a public report alleges that a component on its fleet is misbehaving. That second rule is the one that BLU’s customers would have needed in November 2016. It should name who reads the report, who decides whether devices are pulled, and who tells users what to do.
Who owns the standard
In a large organization, devices pass through several hands. Procurement negotiates the purchase. The mobility or end-user computing team configures and enrolls the devices. Security sets policy and monitors threats. Privacy reviews data flows. Third-party risk management runs vendor intake. Each function can reasonably assume that another one looked at the brand’s preinstalled software. The written standard is what assigns that look to someone.
A workable standard, as we see it, names one owner for supplier oversight of each device line and states the evidence that owner collects. It sets a review frequency tied to risk, so that devices issued to staff handling customer or patient data get more attention than kiosk tablets. It defines the trigger events that force an unscheduled review: a public vulnerability report, a change in the brand’s ownership or manufacturing, or a new preinstalled component pushed by update. And it says who can decide to suspend a device line, because that decision has operational costs that procurement alone cannot weigh.
Without that assignment, the organization’s posture in November 2016 would have depended on whether the right person happened to read the news.
Contract: an agreement that bought updates and said nothing about data
The third failure was “failing to contractually require their service providers to adopt and implement data security standards, policies, procedures or practices.” The complaint did not reproduce the BLU and ADUPS agreement. It described its purpose: firmware updates and nothing else. It alleged that BLU failed to require its service providers, by contract, to adopt data security standards.
An agreement that defines the service but not the data obligations leaves the supplier’s data use to the supplier. BLU’s privacy policy told consumers that service providers could use personal information only for the services they performed. That statement could be true only if the contracts made it true. The complaint’s Count I turned on the difference. ADUPS received contacts and message contents that it did not need for updates, the complaint said, so the policy’s description of limited sharing was false.
The point for any organization is that a privacy notice is a promise about other companies’ behavior. Each sentence describing what service providers may do should map to a clause in a contract that requires it. Where no clause exists, the notice describes a hope. The BLU order’s security program provision makes the link explicit. It requires the company to take reasonable steps to select and retain capable service providers and to require them, “by contract, to implement and maintain appropriate safeguards.”
Clauses that would have mattered
A buyer’s contract with a device brand or reseller cannot reach the brand’s update supplier directly. It can still require the brand to carry certain obligations down its chain. Five kinds of clause follow from the complaint’s facts. A disclosure clause would require the brand to list privileged preinstalled software and its suppliers, and to update the list when it changes. A data-use clause would require the brand to bind its suppliers to collect only what their function needs. A notice clause would require the brand to tell the buyer promptly of any finding or credible report that a component collects data beyond its stated purpose. A remediation clause would commit the brand to fix or remove such a component without cutting off security updates. A support clause would state how long the brand will supply updates for each model.
Not every buyer can negotiate all five. A small credit union buying a few dozen phones through a retailer has little bargaining power. A state agency buying through a statewide contract may have more. The clauses are still useful as a checklist for what the buyer is accepting when it cannot get them. A buyer that knows it has no disclosure clause knows it is relying on the brand’s goodwill and on public research.
The Safeguards Rule had used the same words
That language was not new in 2018. The FTC’s Standards for Safeguarding Customer Information, the Safeguards Rule at 16 CFR part 314, applied to financial institutions under the FTC’s jurisdiction. As it read on May 1, 2018, section 314.4(d) required an institution to oversee service providers in two steps. It had to take reasonable steps to select capable ones, and to be “requiring your service providers by contract to implement and maintain such safeguards.” The BLU order applied that two-step formula to a device seller. Selection and contract were already the standard for non-bank financial firms. The order made them the standard for a phone brand as well.
For banks supervised by the federal banking agencies, the parallel duties sat in those agencies’ own guidance and the interagency safeguards guidelines. We do not attempt to map every regime here. The common thread is that contract terms are evidence. An organization that cannot produce the clause that binds a supplier cannot show that the supplier was bound.
Assessment: nobody evaluated the code that ran with root access
The fourth failure was “failing to adequately assess the privacy and security risks of third-party software, such as ADUPS.” This one concerns the product, not the relationship. The complaint said the preinstalled software contained commonly known vulnerabilities that exposed devices to command injection. An attacker could use them, it said, to gain full access, factory reset a device, capture screenshots and video of the screen, and install malicious applications.
The point here is that two different harms came through the same component. One was deliberate collection by the supplier. The other was a weakness any attacker could use. A selection review and a contract address the first. Only an assessment of the software itself addresses the second. A supplier can be honest and its code still be exploitable.
Assessment also caught a trade-off the complaint described plainly. Consumers who disabled the ADUPS software to stop the collection lost the channel that delivered security updates. The same component was both the privacy risk and the patch path. Removing it closed one exposure and opened another. That trade-off belongs in any assessment of update software, because switching off a misbehaving updater also switches off the patches it carries.
What assessing an update channel means
For an organization with a device fleet, assessing the update channel has three pieces. The first is traffic: what the device sends, to which endpoints, when, and whether that matches what the brand described. The second is privileges: which preinstalled components run with system rights and whether they can be disabled without losing patches. The third is continuity: what happens to the fleet if a component has to be removed or a supplier exits. Each piece produces a record that can be compared with the brand’s statements and with the organization’s own policy.
Assessment after a public report
The complaint described what happened after reports surfaced in November 2016. BLU posted a notice that ADUPS had updated its software. The complaint alleged that on older devices the supplier kept operating without adequate oversight. For a buyer, the lesson of that sequence concerns verification. A supplier’s statement that it has stopped a practice is a claim. An assessment tests the claim on the buyer’s own devices, across the models it actually holds.
Verification needs a baseline. An organization that had recorded which endpoints its devices contacted before the report could compare traffic after the fix. One without a baseline could only accept the brand’s word. This is the practical reason to assess a component before anything goes wrong: the first assessment is what makes the second one meaningful.
Why the update channel concentrated all four failures
The four failures are separate duties, but in the BLU complaint they met in one component. In our reading, that was not an accident of this case. Update software has three properties that make it a place where supplier oversight can fail on every count at once.
It needs high privileges. To install patches and operating system upgrades, it must be able to change the device beneath the user and beneath ordinary applications. Those privileges are what let it read messages, location and call logs if its code chooses to.
It is invisible to the user and, unless the buyer asks, to the buyer. A banking app or an email client shows up on the home screen and in procurement lists. An update client runs in the background, preinstalled, and is described, if at all, as part of the operating system. The buyer’s inventory names the device model, not the firmware supplier.
It cannot simply be removed. As the complaint said, consumers who disabled the ADUPS software lost critical security updates. A component that is both a risk and a control is hard to govern by exclusion. It has to be governed by the four steps the complaint listed, because the alternative of turning it off creates a different exposure.
Those three properties explain why a buyer should look first at the update path when assessing a device line. Applications installed later usually pass through app stores, permission prompts and mobile device management. A preinstalled updater can sit outside those gates.
What the case did not decide
The BLU matter was a settlement. The complaint stated allegations, and the order recorded that the respondents neither admitted nor denied them except for jurisdictional facts. No court found that BLU or ADUPS acted as alleged. We rely on the complaint for what the Commission alleged, and we describe it in those terms.
The legal theory was deception. Both counts turned on statements BLU had made in its privacy policy. The complaint did not allege that collecting the data was unfair in itself, and the proposed order required no payment. In our reading, that means the case does not establish a freestanding duty, for every device seller, to perform the four oversight steps. It establishes that a seller who promises appropriate security and limited sharing can be held to those promises, and that the four steps are what the Commission looked for when testing them. Organizations should ask counsel which duties apply to their own statements.
The case also said nothing directly about buyers. Its order binds the respondents. Everything we say about banks, health systems and agencies is our application of its structure to their position, not a reading of any duty the order placed on them.
What the order added: consent, assessments and records
The proposed order went beyond the four failures. Three provisions stand out for a reader in a regulated organization.
First, it defined “Covered Information” narrowly, as precise geolocation and the content of text messages, audio conversations, photographs or video communications. Before collecting or disclosing such information, the respondents would have to disclose the categories, the identity of any third parties that receive it and every purpose. The disclosure had to be separate from any privacy policy or terms of use, and the consumer had to give affirmative express consent. The order treated location and communications content as a class of data that could not be collected through a general notice.
Second, the order required assessments by a qualified independent professional. It named the qualifications. One route was a Certified Secure Software Lifecycle Professional with experience programming secure consumer-grade devices. Another was a Certified Information System Security Professional with software development security experience and such programming experience, or someone the FTC approved. Each assessment had to describe the safeguards, explain why they fit the company, and certify that the program was operating with sufficient effectiveness. The order tied the assessor’s skills to the product, not to a generic audit role.
Third, the order required records. For 20 years, the respondents had to create records such as complaints, compliance evidence and every widely disseminated statement about how they protect personal information. Retention ran five years for each record. The order kept the promise and the evidence side by side, which is exactly the comparison Count II had to make without them.
The order bound only the respondents, not any buyer of BLU phones. In our reading, its provisions still offer a usable template. A buyer can ask a device brand whether its program has the seven elements the order listed and whether an assessor with device-specific skills has reviewed it. It can also ask whether the brand separates consent for location and message content from its general privacy policy.
The consent provision deserves a separate note, because it reaches past phones. It required disclosure of what is collected, who receives it and why, “separate and apart from” any privacy policy or terms of use, followed by affirmative express consent. Banks, health systems and agencies publish their own mobile apps, and those apps may collect location for fraud checks, appointment check-in or field work. The order did not apply to them. Its structure still shows how the Commission treated location and message content in 2018: as data whose collection needed its own notice and its own yes. An organization reviewing its own apps can ask whether its collection of those categories would meet that bar, and record its answer.
The assessment provision also had a reporting rhythm that matters to buyers. The initial assessment covered the first 180 days after the order issued and had to be submitted to the Commission. Later biennial assessments had to be kept and produced on request. The order gave buyers no right to those reports. It can ask any device brand whether it is subject to an order or audit regime and what its latest independent assessment covered.
How the four steps fall on banks, health systems and agencies
The duties that apply to a buyer differ by sector, and organizations should confirm with counsel which apply to them. The supplier chain, however, looks the same in each.
Banks and other financial institutions
A financial institution’s mobile devices carry customer conversations, one-time passcodes and access to internal systems. A device that sends message contents to an unknown server could expose customer information. Supervisors expect institutions to manage third-party relationships across their life cycle. A device brand is a third party, and in our reading so are the brand’s own suppliers to the extent they touch the institution’s data. Non-bank financial firms under the FTC’s Safeguards Rule had the select-and-contract duty written in the rule itself.
The practical step is to treat device procurement as third-party risk management, not as a hardware purchase. That puts the brand through the same intake as a software vendor, with a questionnaire covering preinstalled components and their data flows.
The size and type of institution changes how this works. A broker-dealer with duties to preserve business communications has, as we see it, an added reason to know what leaves its devices. A small credit union may rely on a single managed service provider to buy and enroll its phones. In that case the provider’s device selection is part of the credit union’s third-party risk, and the questionnaire belongs in the provider’s contract.
Health systems
For a HIPAA covered entity, the Security Rule requires a risk analysis of electronic protected health information. In our reading, clinicians’ phones that receive patient messages, photos or call logs are inside that analysis. A component that copies message contents off the device is a disclosure path. Whether a device brand or its update supplier is a business associate is a legal question that turns on the facts. The BLU facts suggest that, whatever the answer, the risk analysis has to include them.
Health systems also face the removal trade-off acutely. Clinical devices and shared phones are hard to replace quickly. Disabling an update component can leave a fleet unpatched. A plan that decides in advance how to isolate or replace affected devices avoids having to choose between privacy and patching in the middle of an incident.
Ambulatory networks and physician groups face a different version of the problem. They may let clinicians use personal phones for secure messaging, with the organization controlling only an application. Their risk analysis then has to account for devices they did not choose, from brands they cannot question. The four steps shrink to what they can control: approved device lists, minimum patch levels and a rule for when a device model is no longer allowed.
Public agencies
Agencies buy devices through contracts that may pass through several resellers. Federal agencies work under supply chain risk management guidance from NIST, among other requirements; state and local agencies vary. The BLU case adds a concrete test to those frameworks: does the contract give the agency the right to receive, from the device brand, a list of preinstalled privileged software and its suppliers. If not, the agency cannot perform the first of the four steps at all.
Agencies also face a bring-your-own-device question. Where staff use personal phones for work, the agency did not choose the brand. Its controls then sit in mobile device management and application containers. Those controls protect agency data on the device. They may not stop a privileged system component from reading what the device displays.
Counties and cities add a resourcing question. A small jurisdiction may have no third-party risk function at all. For such an agency, the four steps can be carried by a cooperative purchasing contract or a state program that vets device lines once for several buyers. That arrangement only works if the vetting covers preinstalled software and the update path, and if the results are shared with the buyers who rely on them.
Objections: scale, cost and who can inspect firmware
Three objections deserve an answer.
The first is scale. A bank or agency may buy thousands of devices from a few brands. It cannot audit the supply chain behind each component. That is true, and the four steps do not require it. They require asking the brand for the facts, recording the answers and choosing brands that answer. The burden of supplier oversight sits with the brand, as the BLU order placed it. The buyer’s job is to confirm that the brand carries it.
The second is cost. Lower-priced devices are attractive for large fleets, and a stricter intake may push buyers toward more expensive brands. The BLU complaint does not say that price caused the failures, and we draw no such conclusion. What it shows is that the cost of a device includes the cost of knowing what runs on it. A device whose brand cannot describe its preinstalled software carries a cost the price tag omits.
The third is technical. Few organizations can inspect firmware, and traffic analysis may miss collection that is encrypted or infrequent. Text messages here went out every 72 hours. A short test could have missed them. The answer is that assessment does not have to be perfect to be useful. It sets a baseline against which later reports can be checked, and it identifies which components to watch. Public reports played that role in 2016. An organization that knew which of its devices ran the component would have been able to act on the reports quickly.
The objections do not change the structure. Each of the four steps addresses a different way the chain can fail. Skipping one leaves that failure unguarded.
What a procurement and security leader could check
The complaint’s four failures translate into checks that need no new budget line, only an owner.
- For each device line in the fleet, obtain from the brand a list of preinstalled components that run with system privileges, and the company behind each.
- Record what each privileged component sends off the device, to which endpoints, and why, and compare it with what the brand’s privacy statements say.
- Write down who owns supplier oversight for devices, on what schedule evidence is requested, and what happens when a public report names a component in the fleet.
- Check that procurement contracts with device brands and resellers require disclosure of preinstalled software and notice of material changes to it.
- Decide in advance how the organization would keep patching a fleet if an update component had to be disabled or removed.
Each check produces a record. Taken together, they show an examiner or auditor that the organization treated the update channel as a supplier relationship, with selection, standards, contract and assessment, rather than as a feature of the hardware.
With hindsight
The FTC finalized the BLU order on September 10, 2018. Its announcement said the Commission voted 5-0 to give final approval and to respond to six commenters. It restated the order’s core terms. They were no misrepresentations about privacy and security, a comprehensive security program covering risks tied to new and existing devices, and third-party assessments every two years for 20 years. Record-keeping and compliance monitoring completed the list.
The two-step formula for service providers that the order shared with the Safeguards Rule did not stay at two steps. As the rule read in 2018, section 314.4(d) asked financial institutions to select capable service providers and to require safeguards by contract. As checked on November 15, 2026, the same duty sits in section 314.4(f), with a third step: “Periodically assessing your service providers based on the risk they present and the continued adequacy of their safeguards.” The eCFR credits the current section to amendments published on December 9, 2021, and November 13, 2023. The rule also now requires a designated “Qualified Individual,” who may be employed by a service provider. In that case the institution must “Retain responsibility for compliance.” In our reading, the amended rule addresses the gap the BLU complaint described after November 2016: oversight that ends at signing.
Banking supervisors moved in the same direction. On June 6, 2023, the Federal Reserve, the FDIC and the OCC issued final joint guidance on managing risks from third-party relationships. The agencies said it covered risk management “for the stages in the life cycle of third-party relationships” and replaced each agency’s existing general third-party guidance. Device brands and the suppliers behind them are third-party relationships in that sense, although the guidance does not single them out.
The added assessment step changes what a supplier file looks like. Under the 2018 text, an institution could satisfy the service provider duty with a diligence record at selection and a signed contract. Under the current text, it also needs evidence that it reassessed each provider at intervals set by the provider’s risk. For a device brand or a managed mobility provider, that means a dated record of what was asked, what was answered and what changed since the last review. The rule applies to financial institutions under the FTC’s jurisdiction, not to banks supervised by the federal banking agencies, which work under their own guidance.
For federal agencies and their suppliers, NIST published Special Publication 800-161 Revision 1, Cybersecurity Supply Chain Risk Management Practices for Systems and Organizations, in May 2022, with updates as of November 1, 2024. Its abstract describes the concern that acquired products and services may contain malicious functionality, be counterfeit or be vulnerable because of poor manufacturing and development practices. It attributes the risk in part to limited visibility into how acquired technology is developed, integrated and deployed. That limited visibility is the condition the BLU complaint described inside a single phone.
What held up, in our reading, is the structure. Selection, contract and continuing assessment of suppliers recur in each of these later texts; the banking guidance names due diligence and selection, contract negotiation and ongoing monitoring as separate stages, and the Safeguards Rule added periodic reassessment. What did not change is the buyer’s position. No later rule we reviewed gave a bank, health system or agency a direct contract with the company that writes the update software on its phones. The buyer’s lever is still the brand, and the brand’s willingness to say what it installed.
Spotted an error? Tell us and we will check it and correct the article where needed.