FTC staff said in 2016 there is no such thing as NIST Framework compliance; a 2021 rule declined a safe harbor
On August 31, 2016, FTC staff wrote that there is no such thing as complying with the NIST Cybersecurity Framework, and mapped FTC data security cases onto its five functions. In 2021 the Commission declined to make outside frameworks a Safeguards Rule safe harbor.
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
On August 31, 2016, FTC staff wrote that there is no such thing as complying with the NIST Cybersecurity Framework, and mapped FTC data security cases onto its five functions. In 2021 the Commission declined to make outside frameworks a Safeguards Rule safe harbor.
- The week The post presented the perspective of FTC staff.
- The question: why “complying with the Framework” missed the test The FTC post repeated that point.
- The mapping: enforcement allegations matched to five functions It described two more cases.
- The tool: what the Framework offered and what it left to the company That is a governance benefit.
- The proof: records that show a reasonable process The FTC said its cases ask whether a company undertook a reasonable process. [1] A process leaves traces.
- How the post applied to different organizations Non-bank financial companies fall under the FTC’s Safeguards Rule.
- Objections, in order The post said the Framework was consistent with the FTC’s approach. [1] It also said there is no such thing as complying with it. [1] Both statements stand.
- Checks a security or compliance leader could run today Search public statements, contracts and marketing for claims of “compliance” with the NIST Framework.
- With hindsight The Commission returned to the question in rulemaking.
The week
On August 31, 2016, the Federal Trade Commission published a post on its Business Blog titled The NIST Cybersecurity Framework and the FTC. [1] The same day, the agency issued a short release pointing businesses to it. [2]
The release said the post provided guidance on how the framework created by the National Institute of Standards and Technology aligns with the FTC’s data security program. [2] It said the post notes the framework is not a checklist but a method for a company to identify risks and adjust its security efforts. [2] The release called that consistent with the FTC’s focus on reasonable data security. [2]
The post presented the perspective of FTC staff. It opened with a question it said the agency was asked: whether complying with the NIST Cybersecurity Framework meant complying with what the FTC requires. [1]
The post gave the staff’s view up front. It said the Framework was consistent with the process-based approach the FTC had followed since the late 1990s, with the 60-plus law enforcement actions the FTC had brought, and with its educational messages to companies, including its Start with Security guidance. [1]
The post then gave background. It said Executive Order 13636, issued in February 2013, called on NIST to develop a voluntary, risk-based cybersecurity framework for critical infrastructure. [1] NIST issued the Framework in February 2014. [1] The Framework document itself is dated February 12, 2014, and calls itself Version 1.0. [3]
The post described the Framework’s Core. It said the Core is composed of five concurrent and continuous functions: Identify, Protect, Detect, Respond and Recover. [1] Each function is divided into categories, and each category into subcategories pointing to informative references in standards and guidelines. [1]
It gave an example. The post said the Identify function has five categories: Asset Management, Business Environment, Governance, Risk Assessment and Risk Management Strategy. [1] Under Governance, one subcategory calls for an organizational security policy, with references to standards such as COBIT, ISA, ISO/IEC and NIST SP 800-53 Revision 4. [1]
The post then answered its question. It said the Framework is not, and is not intended to be, a standard or checklist. [1] It said the Framework does not include specific requirements or elements. [1] In that respect, it said, there is really no such thing as complying with the Framework. [1]
The post described the FTC’s side. It said Section 5 of the FTC Act is the primary enforcement tool the FTC relies on against deceptive and unfair practices in data security. [1] It said that since 2001 the FTC had settled some 60 cases against companies it alleged failed to provide reasonable protections for consumers’ personal information. [1]
It said the touchstone of the FTC’s approach had been reasonableness. [1] A company’s measures, the post said, must be reasonable in light of the volume and sensitivity of the information it holds, the size and complexity of its operations, the cost of available tools and other factors. [1] It said the FTC’s cases focus on whether a company undertook a reasonable process to secure data. [1]
The post then walked through the five functions. For each, it described FTC complaints whose alleged lapses it said corresponded with the Framework’s guidance. [1] The release summarized this as highlighting enforcement cases in which the security problems alleged mirrored concerns the Framework addresses. [2]
The post closed with advice. It said the Core functions can serve as a model for companies of all sizes to establish or improve a data security program, review current practices, or communicate requirements to stakeholders. [1] It recommended that companies review the FTC’s Start with Security publication alongside the Framework. [1]
In our view, the line that stands out in the FTC’s August 2016 post is that there is really no such thing as complying with the NIST Cybersecurity Framework. [1] The sentence was not a dismissal. The same post said the Framework was consistent with the FTC’s process-based approach to data security. [1]
Our argument is that the post told companies what kind of evidence the FTC would look for. It would not accept a Framework label as proof. It would look at whether a company had followed a reasonable process to identify and address risks to consumer data. [1] When the Commission rewrote its Safeguards Rule in 2021, it declined to make the Framework a safe harbor and instead named specific controls. [4] In our reading, that confirmed the 2016 post’s line: frameworks organize the work, and the regulator judges the work.
We take the post in four parts:
- The question: why “complying with the Framework” was the wrong question, and what the FTC’s reasonableness test asked instead.
- The mapping: how the post matched FTC enforcement allegations to the Framework’s five functions.
- The tool: what the Framework offered a company, and what it left to the company.
- The proof: what records would show a reasonable process to a regulator reading after the fact.
The parts do not overlap. The first is the legal standard, the second the enforcement record, the third the Framework itself and the fourth the evidence connecting them. Together they cover what the post said and what it implied for a security program. After them we look at how it applied to different organizations, take the objections in order, and end with checks a security or compliance leader could run.
The question: why “complying with the Framework” missed the test
What the Framework says about itself
The Framework described itself as a set of industry standards and best practices to help organizations manage cybersecurity risks, created through collaboration between government and the private sector. [3] It said it was not a one-size-fits-all approach. [3] Organizations would continue to have unique risks, it said, and how they implemented its practices would vary. [3]
The FTC post repeated that point. It said the Framework is meant to help an organization determine its current capabilities, set goals and plan improvements, but does not include specific requirements or elements. [1] A document with no requirements cannot be complied with. It can only be used.
What the FTC asked instead
The post said the FTC’s touchstone was reasonableness. [1] It listed the factors: the volume and sensitivity of information, the size and complexity of operations, the cost of available tools to address vulnerabilities, and other factors. [1]
It also said the FTC had recognized from the outset that there is no such thing as perfect security. [1] Security, the post said, is a continuing process of detecting risks and adjusting a program and defenses. [1] The FTC’s cases focus on whether the company has undertaken a reasonable process to secure data. [1]
In our reading, that is a test of conduct over time. A company is judged on what it knew, what it did and how it adjusted. A framework can structure that conduct. It cannot substitute for it.
Process, not outcome
The post’s test is about process. [1] A breach does not prove a process was unreasonable, and the absence of a breach does not prove it was reasonable. The FTC said it recognized there is no perfect security. [1]
That framing protects a careful company that is breached anyway. It does so only if the company can show what it did and why. A program that existed but left no trace looks, after a breach, much like a program that did not exist. It also exposes a careless one that has been lucky. In our reading, the practical consequence is that the record of the process matters before anything goes wrong, because that record is what a regulator reads afterward.
Why the question kept being asked
The post said the agency was asked the compliance question. [1] The reason is not hard to see. A company facing an uncertain legal standard would like a document it can point to. The Framework, issued by a federal agency, looked like such a document.
The post declined to give it that role. It described the Framework as fully consistent with the FTC’s approach, because the things the Framework asks organizations to evaluate are the things the FTC had been evaluating in its Section 5 cases. [1] Consistency is not equivalence. Two documents can ask the same questions without either one supplying the answers.
The risk of treating a label as proof
A company that tells customers or partners it “complies with the NIST Framework” claims something FTC staff said there is really no such thing as. [1] If the underlying practices fall short, the claim can make matters worse. A deception theory under Section 5 turns on what a company represented. The post itself did not discuss that risk, so we flag it as our reading, not the FTC’s statement.
Reasonable in light of what
Each of the post’s factors points to something a company can document. [1] Volume and sensitivity of information point to a data inventory. Size and complexity point to an organizational description. Cost of tools points to a record of options considered.
The phrase “and other factors” leaves room for the agency. [1] In our reading, it also leaves room for the company to explain circumstances specific to it, such as a recent acquisition or a legacy system awaiting replacement, provided the explanation was written down at the time.
What the FTC had settled
The post said that since 2001 the FTC had settled some 60 cases against companies it alleged failed to provide reasonable protections. [1] Settlements are not adjudications. The orders bind the parties; the complaints describe allegations.
The post used those allegations as teaching material. That is a fair use of the record, as long as readers remember what it is. In our reading, a company should treat each alleged lapse as a control question to answer for itself, not as a finding about what the law requires in every case.
The mapping: enforcement allegations matched to five functions
Identify
The post said the Identify function helps organizations understand cybersecurity risks to systems, assets, data and capabilities. [1] It described FTC complaints alleging companies failed to implement policies and procedures to safeguard consumer information. [1] It said those allegations align with the Framework’s guidance on an organizational information security policy. [1]
It also described complaints alleging companies had no process for receiving, addressing or monitoring reports about security vulnerabilities. [1] It linked those to the Framework’s guidance on receiving threat and vulnerability information from information-sharing forums and sources. [1]
Protect
For Protect, the post described a complaint alleging a company gave almost all its employees administrative control over its system. [1] According to the complaint, as the post described it, that increased the risk that compromise of any employee’s credentials could cause a serious breach. [1] The post tied this to the Framework’s guidance on access permissions, least privilege and separation of duties. [1]
It described two more cases. In one, the FTC alleged an employee transported a laptop with personal information in a way that made it vulnerable to theft. [1] In the other, the FTC alleged a company transported portable media with personal information in a way that made it susceptible to theft. [1] The post said that in both cases the laptops and media were stolen. [1]
The post linked these to the Framework’s guidance on protecting data in transit and managing assets through removal, transfer and disposition. [1]
Detect
For Detect, the post described a complaint alleging a company did not use an intrusion detection system and did not monitor system logs for suspicious activity. [1] It described another alleging a company did not inspect outgoing internet transmissions to identify unauthorized disclosures of personal information. [1]
The post said that had these companies used tools to monitor activity on their networks, they could have reduced the risk or breadth of a compromise. [1] It said their alleged deficiencies did not comport with the Framework’s guidance on monitoring networks for events and for unauthorized personnel, connections, devices and software. [1]
Respond
For Respond, the post described a case in which the FTC alleged a company failed to follow proper incident response procedures, including failing to monitor its network for malware used in a previous intrusion. [1] As the post recounted the allegations, intruders accessed the network on three separate occasions in a 21-month period. [1]
The post said, again describing the FTC’s allegations, that the intrusions led to compromise of more than 619,000 payment card account numbers and more than $10.6 million in fraud loss. [1]
It described another case in which, according to the complaint, a device maker learned of vulnerabilities in its routers but failed to give consumers adequate notice of the risks, mitigations and available updates. [1] As a result, the post said in describing the complaint, attackers gained unauthorized access to over 12,900 connected storage devices. [1] It said the case showed the importance of sharing information with external stakeholders. [1]
Recover
For Recover, the post pointed to an FTC order requiring a software company to give broad notice to users about a settlement and how to address vulnerabilities in its software. [1] It said the company reached users through its website and social media and by working with outside parties such as antivirus vendors and browsers. [1]
The post tied that to the Framework’s guidance on communicating recovery activities with internal and external parties. [1] It said FTC orders showed how consumer interests should factor into a company’s recovery plan. [1]
What the mapping showed
The post’s examples were allegations and orders, not court findings, and the post described them that way. [1] Read together, they show the FTC’s cases covered all five functions. A company that thought of data security as prevention alone would miss three of the five.
In our reading, the mapping also shows where a regulator’s attention falls after an incident. Detect and Respond allegations look backward from the breach. They ask whether the company could have seen the intrusion sooner and contained it better. Those are questions answered by logs, alerts and incident records.
The allegations as a control list
Read as a list, the post’s examples name concrete controls. [1] A written security policy. A channel for receiving vulnerability reports. Least privilege for administrative access. Protection for laptops and media in transit. Intrusion detection and log review. Inspection of outbound data. Monitoring for malware from a past intrusion. Notice to users about vulnerabilities and fixes.
None of those items is exotic. Each corresponds to one or more Framework subcategories, as the post said. [1] A company could check its program against the list in an afternoon and learn where it stood on the issues the FTC had already raised.
Where the examples came from
Judging by the companies the post named, in our reading, the examples spanned retail, health services, technology products and services, and hospitality. [1] The common thread was not the industry but the data: payment cards, health information and personal information held by companies that had promised, or were expected, to protect it.
That breadth is why the post reads as cross-sector guidance. The same controls appear whether the company sells pet supplies or routers. [1]
The tool: what the Framework offered and what it left to the company
A common language
The post said the Framework can deliver benefits beyond security itself. [1] It said the Core provides a common language that can help discussions between IT staff and business people, some of whom may tune out technical terminology. [1]
That is a governance benefit. A board that hears about five functions and a handful of categories can follow a security report better than one that hears about tools. The Framework’s structure lets a company report on its program in terms a non-specialist can track.
Informative references
The Framework’s subcategories point to informative references: specific sections of standards and guidelines that illustrate how to achieve each outcome. [1] The post’s example was the Governance subcategory on security policy, pointing to COBIT, ISA, ISO/IEC and NIST SP 800-53. [1]
The references are examples, not mandates. A company chooses which to use. The choice, and the reason for it, is part of the process the FTC said it would examine. [1]
Profiles and Tiers
The Framework had three parts: the Core, Profiles and Implementation Tiers. [3] The Framework said Profiles help an organization align its cybersecurity activities with business requirements, risk tolerances and resources. [3]
A current Profile describes where a company is. A target Profile describes where it wants to be. The gap between them is a plan. In our reading, that plan, with dates and owners, is the kind of document that shows a reasonable process.
Implementation Tiers
The Framework described four Tiers, from Partial to Adaptive. [3] It said Tiers describe the degree to which an organization’s risk management practices show the characteristics the Framework defines, such as being risk-aware, repeatable and adaptive. [3]
The Framework also said Tiers do not represent maturity levels. [3] It said successful implementation rests on achieving the outcomes in the organization’s Target Profile, not on a Tier determination. [3] That is another reason a Tier label, like a compliance label, would not answer the FTC’s question.
A seven-step approach
The Framework set out steps for establishing or improving a program, beginning with prioritizing and scoping, then orienting by identifying systems, assets, regulatory requirements and threats. [3] The later steps include creating a Current Profile, conducting a risk assessment, creating a Target Profile, analyzing gaps and implementing an action plan. [3]
Each step produces an artifact. In our reading, those artifacts are the record the FTC post implied a company should be able to show: the scope, the risks, the gaps and the plan.
What the Framework left open
The Framework does not say how much encryption is enough or how often to test. It says organizations will implement practices differently. [3] The FTC’s reasonableness factors, such as sensitivity of data and cost of tools, fill that gap case by case. [1]
So a company using the Framework still has to make the decisions the FTC would judge. The Framework helps it ask the right questions. It does not answer them.
Version 2.0 asks the same question
NIST’s 2024 revision keeps the same stance on prescription. It says the Framework offers a taxonomy of high-level cybersecurity outcomes and does not prescribe how outcomes should be achieved. [6] It links instead to resources on practices and controls that could be used. [6]
Its stated audience goes beyond those who run security programs. It names executives, boards of directors, acquisition professionals, lawyers, risk managers and auditors alongside those leading security programs. [6] A Framework addressed to lawyers and auditors invites the compliance question again. The 2016 post’s answer still fits it: the Framework describes outcomes, and a regulator judges how a company pursued them. [1] [6]
Start with Security as the companion
The post recommended pairing the Framework with the FTC’s Start with Security guidance. [1] It described that publication as summarizing lessons from the FTC’s data security cases and giving practical guidance. [1]
The pairing is telling. The Framework is general. The FTC’s guidance is drawn from specific allegations. A company that used both would have a structure and a list of things that had gone wrong for others.
The proof: records that show a reasonable process
A written risk assessment
The FTC said its cases ask whether a company undertook a reasonable process. [1] A process leaves traces. The first is a risk assessment that names the information held, where it is and what could go wrong.
The Framework’s Identify function, with its asset management and risk assessment categories, points to the same document. [1] A company that keeps a dated assessment, updated when its business or threats change, has the starting point of the record.
Decisions with reasons
Reasonableness turns on cost and sensitivity. [1] So a record that shows why a control was chosen, or why one was not, matters as much as the control itself. A decision to delay encryption on a system, with the reason and a compensating control, is evidence of process. An undocumented gap is not.
Monitoring that produces records
The Detect allegations concerned missing intrusion detection and unmonitored logs. [1] A company that monitors produces records: alerts, reviews and escalations. Those records show monitoring happened. Without them, a company cannot show it would have noticed an intrusion.
Incident records
The Respond allegations included failing to monitor for malware from a previous intrusion. [1] An incident record that shows what was found, what was fixed and what was checked afterward is the evidence that response was more than containment.
Board and management reporting
The post said the Framework’s common language helps conversations between IT staff and business people. [1] A written report to the board or senior management turns those conversations into a record.
For firms under the Safeguards Rule, that report is now required. The current text requires the program’s Qualified Individual to report in writing, regularly and at least annually, to the board or equivalent body, or to a senior officer if there is none. [5] The report must cover the program’s overall status and compliance with the rule, and material matters such as risk assessment and risk management decisions. [5]
Vulnerability intake records
The Identify examples included companies alleged to have no process for receiving vulnerability reports. [1] A company that has such a process keeps records: reports received, triage decisions, fixes and notices. Those records show the channel works.
Training records
The post said the Protect function includes raising employees’ awareness and providing training. [1] Training leaves records: who was trained, on what and when. The Safeguards Rule now requires covered firms to give personnel security awareness training updated to reflect risks found in the risk assessment. [5]
A training record linked to the risk assessment shows the program responded to the risks it found. A generic annual module with no link to identified risks shows less.
A written incident response plan
The Respond examples concerned failures in incident response. [1] The Safeguards Rule now requires covered firms to keep a written incident response plan designed to promptly respond to and recover from security events materially affecting customer information. [5]
The rule lists what the plan must address, beginning with its goals, the internal processes for responding and the definition of clear roles and responsibilities. [5] For firms outside the rule, the same headings make a reasonable outline. The plan, and records of its use, show the process existed before the incident.
Vendor and partner communications
The Recover example involved communicating with users and outside parties. [1] A company’s record of whom it notified, when and with what instructions shows it treated recovery as including the people affected.
Evidence of adjustment
The post described security as a continuing process of detecting risks and adjusting a program and defenses. [1] Adjustment leaves its own trace: a change made because a test, an incident or a business change revealed a gap.
The Safeguards Rule later wrote this into its text for covered firms. It requires them to evaluate and adjust their programs in light of testing and monitoring results and material changes to operations or business arrangements. [5] A change log that ties each adjustment to its trigger is the clearest record that the process was continuing rather than a one-time project.
How the post applied to different organizations
A non-bank lender or fintech
Non-bank financial companies fall under the FTC’s Safeguards Rule. The rule’s current text says it applies to financial institutions over which the FTC has jurisdiction, naming examples such as mortgage lenders, payday lenders, finance companies, mortgage brokers, account servicers, check cashers and wire transferors. [5] For these firms, the 2016 post described the Section 5 baseline. The Safeguards Rule added specific requirements later. [4]
A bank or credit union
The Safeguards Rule’s scope excludes financial institutions subject to another regulator’s enforcement under section 505 of the Gramm-Leach-Bliley Act. [5] The current text also lists non-federally insured credit unions among the entities it covers. [5] In our reading, that leaves banks and federally insured credit unions to their own regulators’ security standards. The Framework is still a common language for them. The FTC post’s reasoning about process, though, describes the FTC’s test, not a banking agency’s.
A health system or health app
In our reading, a health system that is a HIPAA covered entity answers mainly to HHS for security of patient data, while a consumer health app outside HIPAA is a candidate for Section 5 scrutiny. Which regime applies is a question for counsel. Where the FTC is the regulator, the post’s reasoning applies directly: the question is whether the process was reasonable for the sensitivity of the data. [1]
A software or device maker
The router case in the post concerned a device maker’s handling of known vulnerabilities and its notice to consumers. [1] For any company that ships software or connected devices, the post’s Identify and Respond examples point to two processes: receiving vulnerability reports and telling customers about risks and fixes. [1]
Those processes cross teams. Engineering fixes the flaw. Support and communications tell customers. The record of both is the evidence.
A government contractor
A contractor that handles consumer data for an agency may be bound by contract terms and federal standards. The Framework was built for critical infrastructure under an executive order. [3] A contractor using it for one customer can use the same structure across others, but each contract’s specific requirements still control.
A service provider to financial institutions
A company that processes data for a non-bank lender or broker is a service provider under the Safeguards Rule. The current text requires covered institutions to oversee service providers. [5] They must take reasonable steps to select providers capable of appropriate safeguards, require those safeguards by contract and periodically assess providers based on risk. [5]
For the provider, that means its clients will ask for evidence. A provider that can show a Framework-organized program, with a risk assessment, monitoring records and incident plans, can answer those requests from one body of evidence. The 2016 post’s reasoning about process applies to the provider’s own conduct as well. [1]
A company that builds its own applications
The Safeguards Rule now requires covered firms to adopt secure development practices for in-house applications that transmit, access or store customer information. [5] The post’s Identify examples about vulnerability reports apply to in-house software too. [1]
For a fintech that writes its own code, secure development and vulnerability intake are two halves of one process. The first prevents flaws. The second catches the ones that get through. Both leave records a regulator can read.
A small business
The post said the Core functions can serve as a model for companies of all sizes. [1] The FTC’s reasonableness factors include size and the cost of tools. [1] For a small business, that means a simpler program can be reasonable. The process still needs to exist and be recorded.
Objections, in order
1. If the FTC endorsed the Framework, using it should be enough
The post said the Framework was consistent with the FTC’s approach. [1] It also said there is no such thing as complying with it. [1] Both statements stand. Consistency means the Framework asks the right questions. The FTC judges the answers.
2. Reasonableness is too vague to plan against
The post listed the factors the FTC weighs. [1] It also listed the kinds of lapses the FTC had alleged. [1] Together they give a planning basis: assess risk, address the lapses named in past cases, and document why.
For the firms the Safeguards Rule covers, the Commission later added detail. It said its 2021 amendments were designed to give covered financial institutions more guidance on aspects of their programs, such as access controls, authentication and encryption. [4]
3. A blog post is not law
That is true. The post presented the view of FTC staff. [1] It did not bind the Commission or a court.
But it described the Commission’s enforcement history and the factors it uses. [1] A company reading it learned what the agency had alleged before. That is useful whether or not the post itself has legal force.
4. Mapping cases to the Framework after the fact proves nothing
The post counted cases settled since 2001, and the Framework dates from February 2014. [1] [3] So the post matched allegations to functions retrospectively.
The point of the mapping was not to show the FTC had used the Framework. It was to show the FTC’s concerns and the Framework’s categories overlap. A company that covers the categories is more likely to have addressed what the FTC has alleged.
5. The Framework is for critical infrastructure, not ordinary businesses
The Framework was developed under an executive order on critical infrastructure. [3] The post said that for most organizations, critical infrastructure or not, it may be well worth using for its stated goal of improving risk-based security. [1] NIST’s 2024 version describes itself as usable by any organization regardless of size, sector or maturity. [6]
6. Named controls make frameworks unnecessary
The 2021 Safeguards Rule added specific requirements for covered firms, including encryption, multifactor authentication and board reports. [5] A reader might take that as replacing the Framework.
The Commission did not see it that way. It retained what it called the rule’s process-based approach, letting institutions tailor programs to their size, complexity, operations and data. [4] Named controls set a floor. A framework still organizes everything above it.
7. Regulators should publish a checklist and stop judging process
A checklist would be easier to plan against. The Commission considered something close to it in its Safeguards rulemaking, asking whether to require an outside framework or treat compliance with one as a safe harbor. [4]
It declined both. It said it was not clear detailed frameworks would apply well to institutions of different sizes and industries, and that mandating one would reduce the rule’s flexibility. [4] It said a safe harbor would require it to keep monitoring whether outside standards stayed adequate and to verify compliance with them in investigations. [4] The Commission chose named elements inside a process-based rule instead. [4]
Checks a security or compliance leader could run today
The language test
Search public statements, contracts and marketing for claims of “compliance” with the NIST Framework. Replace them with accurate descriptions of how the Framework is used. FTC staff said there is really no such thing as complying with it. [1]
The risk assessment
Check that a written risk assessment exists, is dated and was updated after the last major change. The Safeguards Rule now requires a written one for covered firms, with criteria for evaluating risks and for how risks will be mitigated or accepted. [5]
The decision log
Check for records of why key controls were chosen and why gaps were accepted, with owners and dates.
Detection evidence
Check that logs are monitored and that alerts, reviews and escalations are recorded. The post’s Detect examples turned on missing monitoring. [1]
Incident follow-through
For the last significant incident, check that the record shows what was fixed and what was checked afterward, including monitoring for the same malware. [1]
Vulnerability intake
Check that the company has a published way to receive vulnerability reports and a record of how each was handled. [1]
The control list from the post
Walk through the post’s examples and record, for each, the control the company uses and the evidence it keeps: policy, vulnerability intake, least privilege, data in transit, intrusion detection, outbound inspection, incident follow-up and user notice. [1]
Service provider evidence
List the service providers that handle customer data. Check that contracts require safeguards and that each provider has been assessed recently, in proportion to its risk. Firms under the Safeguards Rule must do both. [5]
These checks are good practice drawn from the post and later rules. They are not a statement of what the law requires of any given company, which depends on the rules that apply to it and should be confirmed with counsel.
With hindsight
The Commission returned to the question in rulemaking. On December 9, 2021, the FTC published a final rule amending its Standards for Safeguarding Customer Information, the Safeguards Rule. [4] It was effective January 10, 2022, with the provisions in section 314.5 applicable beginning December 9, 2022. [4]
The FTC described five main changes. [4] The rule added guidance on specific aspects of an information security program, such as access controls, authentication and encryption. [4] It added accountability provisions, such as periodic reports to boards. [4]
It exempted institutions holding less customer information from some requirements. [4] It expanded the definition of financial institution to include entities engaged in activities incidental to financial activities, bringing in finders. [4] And it moved several definitions into the rule itself. [4]
The Framework question came up directly. The Commission said its proposal had asked whether the rule should incorporate outside standards, such as the NIST framework, either as required elements or as a safe harbor. [4] It said some commenters favored adopting an outside standard, others favored a safe harbor, and others opposed both. [4]
The Commission declined to incorporate or reference a particular standard or framework. [4] It said it was not clear that detailed frameworks would apply well to institutions of various sizes and industries, and that mandating one would reduce the rule’s flexibility. [4]
It also declined a safe harbor. [4] It said it would have to monitor whether outside standards continued to provide equivalent protection, and in investigations would have to verify independently whether an institution had complied with the outside framework. [4] In our reading, that reasoning restates the 2016 post’s point in rulemaking form: a framework is not a compliance test the agency can simply check. [1] [4]
The Commission said the amendments kept the rule’s process-based approach, letting institutions tailor programs to their size, complexity, operations and the sensitivity of the information they hold. [4] It said, for example, that the new written risk assessment would reflect each institution’s own structure and operations. [4]
As checked on November 16, 2026, the eCFR text of 16 CFR 314.4 lists the required elements. [5] They include a Qualified Individual, a written risk assessment, safeguards including access controls, encryption of customer information in transit over external networks and at rest, and multifactor authentication for any individual accessing any information system. [5]
They also include regular testing or monitoring, with annual penetration testing and vulnerability assessments every six months absent effective continuous monitoring, service provider oversight, a written incident response plan and an annual written report to the board. [5] The section’s amendment history cites the December 9, 2021 rule and an amendment at 88 FR 77508, dated November 13, 2023. [5]
The current text includes a notification duty. A covered institution must notify the FTC of a notification event involving the information of at least 500 consumers as soon as possible, and no later than 30 days after discovery. [5] Section 314.6 exempts institutions holding information on fewer than five thousand consumers from several elements, including the written risk assessment and the board report. [5]
NIST revised the Framework itself. On February 26, 2024, it published the NIST Cybersecurity Framework 2.0. [6] The Core now has six functions: Govern, Identify, Protect, Detect, Respond and Recover. [6] NIST describes Govern as covering the organization’s cybersecurity risk management strategy, expectations and policy, established, communicated and monitored. [6]
The 2.0 document says it can be used by any organization regardless of size, sector or maturity. [6] It keeps the stance the FTC staff relied on in 2016: it does not prescribe how outcomes should be achieved. [6]
ftc.gov refused our automated requests for the post when we checked on November 16, 2026, so we cite the Internet Archive copy of its original address captured September 2, 2016. [1] We make no claim about whether or where ftc.gov still publishes it.
Spotted an error? Tell us and we will check it and correct the article where needed.