The 2017 equities charter put patch uptake among the questions that decide whether a flaw is disclosed
On November 15, 2017, the White House published its charter for deciding when agencies tell vendors about newly found flaws. Its questions included whether enough systems would install a patch. In our reading, that makes defenders’ patch speed an input to the government’s disclosure decision.
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 November 15, 2017, the White House published its charter for deciding when agencies tell vendors about newly found flaws. Its questions included whether enough systems would install a patch. In our reading, that makes defenders’ patch speed an input to the government’s disclosure decision.
- The week On Wednesday, November 15, 2017, the White House published an unclassified charter titled Vulnerabilities Equities Policy and Process for the United States Government.
- The gate: newly discovered, not publicly known The charter applies only to vulnerabilities that are both newly discovered and not publicly known.
- The scale: four kinds of equity, with patch uptake among the defensive ones The charter requires decision-makers to weigh defensive, intelligence and law enforcement, commercial and international equities.
- The clock: business days for decisions, a year for restrictions The charter sets explicit timelines.
- The handoff: after disclosure, the work moves to vendors and defenders The fourth part covers what disclosure sets in motion.
- How this reached different organizations The charter binds federal agencies.
- Objections, in order The argument that the charter made patch uptake a factor in disclosure faces objections.
- What an agency security leader could check Reporting is the part these checks share.
- With hindsight On December 19, 2017, the White House publicly attributed the WannaCry attack to North Korea.
The week
On Wednesday, November 15, 2017, the White House published an unclassified charter titled Vulnerabilities Equities Policy and Process for the United States Government. The charter sets out how federal departments and agencies decide whether to disclose a newly discovered, not publicly known vulnerability to the vendor, or to restrict knowledge of it for a time. It superseded a policy dated February 16, 2010.
The charter states its primary focus plainly. It is to prioritize the public’s interest in cybersecurity and protect core internet infrastructure, critical infrastructure and the economy by disclosing vulnerabilities the government finds. The exception is a demonstrable, overriding interest in using the vulnerability for lawful intelligence, law enforcement or national security purposes. The charter adds that the choice is not always binary.
The options between full disclosure and restriction are listed. The government can share mitigation information with certain entities without disclosing the flaw. It can limit its own use of the vulnerability. It can inform U.S. and allied government entities at a classified level. It can use indirect means to inform the vendor. Each choice is meant to weigh the risks of dissemination against the benefits of government use.
The charter names its governing body, the Equities Review Board. Its members represent the Office of Management and Budget and the Office of the Director of National Intelligence. They also represent the Departments of the Treasury, State, Justice, Homeland Security, Energy, Defense and Commerce, and the Central Intelligence Agency. The board meets monthly. The National Security Agency serves as executive secretariat, and the White House Cybersecurity Coordinator serves as VEP Director.
The process has fixed steps and clocks. An agency that finds a qualifying vulnerability submits it with a recommendation. The secretariat notifies agency points of contact within one business day. Agencies claiming an equity respond within five business days. Disagreements go to discussion within seven business days, then to the board. If disclosure is chosen, dissemination should happen when possible within seven business days. Restricted vulnerabilities are reassessed annually.
Annex B lists the considerations. The defensive questions ask where a product is used, how severe the flaw is, and how likely adversaries are to find it. They also ask whether enough government systems, businesses and consumers will install a patch to offset the harm of educating attackers. Another asks what share of vulnerable systems will remain unpatched for more than a year. Other parts cover operational, commercial and international equities.
The same day, the White House published an article by its Cybersecurity Coordinator explaining the release. The article described the process as a balance between disclosing flaws so vendors can patch them and keeping some for national security or law enforcement use. It said recent cybersecurity incidents of global impact had raised interest in a process that could withstand scrutiny. It said an annual report with metrics would be released.
The article did not name those incidents. One public example from that spring was WannaCry. In May 2017, US-CERT had published indicators for the WannaCry ransomware. Its alert said initial reports indicated the attackers gained access through a critical Windows SMB vulnerability. It noted that Microsoft had released a security update for that vulnerability, MS17-010, on March 14, 2017. It added that Microsoft released patches for older Windows versions on May 13, 2017.
The charter does not name WannaCry or any specific vulnerability. It describes a decision process. It also states, in its background section, that the government did not create the vulnerabilities and that disclosure in the vast majority of cases is clearly in the national interest. Those two statements frame the rest of the document.
The Vulnerabilities Equities Process is read as a national security document, and it is one. It is also, in our reading, a document about patching. Its central defensive questions are about who uses a product, how quickly a fix will arrive and how quickly it will be installed. The government’s willingness to disclose a flaw is tied, in writing, to how defenders behave after disclosure.
That tie has a practical meaning for agency chief information security officers, and for security leaders at banks and hospitals. When the government decides whether to tell a vendor about a flaw, the charter asks it to weigh whether the resulting patch will be applied. Defenders who patch slowly carry their own risk, and more. Their patch speed is one of the facts the charter asks the government to weigh. We take the charter in four parts:
- The gate: which vulnerabilities enter the process and which stay out.
- The scale: the considerations that decide disclosure or restriction, with patch uptake among them.
- The clock: the timelines the charter sets for decisions and dissemination.
- The handoff: what happens after disclosure, when the vendor patches and defenders install.
The parts are distinct. The gate decides whether the process applies at all. The scale decides the outcome. The clock sets how fast. The handoff covers what the outcome requires of everyone outside the process. Together they cover the charter from submission to installed fix. After the four parts, we look at how this reached different organizations, at objections, and at what an agency security leader could check.
A note on our reading. We describe the charter from its own text and the White House article published with it. Annex C, which describes excepted categories, is classified, and we do not speculate about it. Where we draw conclusions about how the charter affects defenders, those are our reading, not statements the charter makes.
The gate: newly discovered, not publicly known
The charter applies only to vulnerabilities that are both newly discovered and not publicly known. Its definitions make the gate narrow. A vulnerability is publicly known if the vendor is aware of it or information about it is in the public domain. A zero-day vulnerability is one unknown to the vendor, exploitable and not publicly known.
What stays outside
Several categories are excluded from the evaluation. Misconfiguration that trades security for availability or ease of use is out. So is misuse of device features, misuse of engineering tools, and the discovery that a device has no security features by design. Vulnerabilities found through security research or incident response that are meant to be disclosed rapidly are also not adjudicated by the process.
The incident response exception matters for defenders. If an agency finds a flaw while responding to an intrusion and intends to disclose it quickly, the process does not slow it down. The charter also says it does not prevent the government from taking immediate action to protect its networks or to warn entities actively threatened by a malicious cyber event.
What a submission contains
The charter sets a minimum for each submission: a description of the vulnerability, identification of the vulnerable products or systems, and a recommendation on dissemination. The secretariat keeps records of every vulnerability submitted, including the submitting agency, the determination and its date, and whether reassessment is needed. Those records exist so the process can be reviewed later.
For an agency, the submission standard is a useful template even outside the process. A finding that names the product, the affected versions and a recommendation is ready for a decision. A finding that lacks them is not. Agencies that adopt the same minimum for internal vulnerability reports make their own triage faster, whether or not a report ever reaches the board.
Classified exceptions
The charter allows specific, limited categories to be excluded from review, listed in the classified Annex C. These are cataloged by the originating agency and reported to the chair of the board, with quantities shared at board meetings. The charter also notes that partner agreements, such as non-disclosure agreements and memoranda of understanding, may restrict disclosure options.
For defenders, the exceptions mean the process is not a complete account of what the government knows. A vulnerability can sit outside it for reasons the public cannot see. That does not make the process empty. It means defenders should not read the existence of the charter as a promise that every government-held flaw is weighed for disclosure.
The records and the annual report
The charter also creates a paper trail around the gate. The secretariat produces an annual report covering October 1 to September 30. It goes to agency points of contact and the National Security Council staff, at the lowest classification possible, with at least an unclassified executive summary. The charter says the report may be provided to Congress, and that it will include statistics the VEP Director considers appropriate.
For defenders, the report promised a way to see the process in aggregate, if not case by case. Whether outsiders would ever see such numbers depended on later choices. The hindsight section describes what Congress required. The charter itself committed only to producing the report and to the possibility of sharing it.
Why the gate is drawn this way
The narrow gate follows from the decision the charter makes. If a vulnerability is already known to the vendor or the public, there is nothing for the government to decide about disclosure. The vendor can fix it and defenders can act. The charter concentrates its process where the government’s knowledge is the only knowledge, which is where its choice matters.
The scope is broad in another direction. The charter covers commercial products, open-source software, government products and industrial control systems, including supervisory control and data acquisition systems. It applies to all government components and personnel, including contractors. A flaw found by a contractor working for an agency enters the same process as one found by agency staff.
Researchers outside government
The gate concerns what the government knows. Independent researchers who find flaws are outside it, and the charter treats their reports as information to safeguard. It says agencies will protect vulnerability information identified by businesses, researchers and foreign governments. It also says the government will encourage them to disclose consistent with international standards or best practices. In our reading, the disclosure decision for such a report stays with the researcher and the vendor.
Contractors and the gate
The contractor point has a procurement consequence. An agency that hires a firm for testing or research may receive vulnerability findings. The charter says its policy applies to contractors. In our reading, agencies should make sure their contracts route such findings to the agency’s VEP point of contact, rather than leaving the contractor to decide what to do with them. Counsel and contracting officers would need to confirm how this applies in a given contract.
The scale: four kinds of equity, with patch uptake among the defensive ones
The charter requires decision-makers to weigh defensive, intelligence and law enforcement, commercial and international equities. Annex B lists the questions under each heading. The list is not exhaustive, the charter says, but it represents concerns that should apply to every decision.
The defensive questions
The defensive questions fall into four groups: threat, vulnerability, impact and mitigation. Threat asks where the product is used, how widely, and whether threat actors would exploit the flaw if they knew of it. Vulnerability asks what access an attacker needs, whether the flaw alone causes harm, and how likely adversaries are to find it. Impact asks how much users rely on the product’s security and how severe exploitation would be.
Two impact questions concern the patch itself. One asks how likely it is that adversaries will reverse engineer a patch, find the flaw and use it against unpatched systems. The other asks whether enough government systems, U.S. businesses and consumers will actually install the patch to offset the harm of educating attackers. Disclosure, in this framing, carries a cost. A patch reveals the flaw, and only installation captures the benefit.
The mitigation questions
The mitigation group asks whether the product can be configured to mitigate the flaw, and whether best-practice guidance already covers it. It asks how likely the vendor is to release an effective patch, how likely that patch is to be applied, and how soon. It asks what share of vulnerable systems will remain unpatched permanently or for more than a year. And it asks whether exploitation can be detected.
Read together, these questions model the world after disclosure. The government is asked to estimate vendor responsiveness and defender behavior. A product whose users patch quickly scores better for disclosure, in our reading, because the window of harm after the patch is shorter. A product whose users patch slowly, or cannot patch, makes disclosure less attractive on these questions, because the patch may teach attackers more than it protects.
The other equities
The intelligence and law enforcement questions ask whether the flaw supports collection, operations or evidence gathering, what its demonstrated and potential value is, and whether alternatives exist. They ask whether disclosure would reveal sources or methods. The commercial question asks what revealing government knowledge of the flaw would mean for relationships with industry. The international question asks the same about foreign relations.
The defensive questions sit alongside these, and the charter does not assign weights. It says decisions should use repeatable techniques where possible, with assessment factors such as prevalence, reliance and severity. Its background section also says that, in the vast majority of cases, responsibly disclosing a newly discovered vulnerability is clearly in the national interest. The weighing happens in a closed process, and the public sees only the framework.
What the patch questions mean for defenders
Our argument rests on the patch questions. They mean that defenders’ behavior is a fact in the government’s decision. Agencies, banks, hospitals and utilities that patch slowly make the defensive case for disclosure weaker on the charter’s own terms. Those that patch quickly make it stronger. The charter does not describe consulting individual defenders, or say how this factor is measured. But the question is written into the policy.
This is a reason for security leaders to treat patch speed as a public good as well as a local control. An agency’s patching performance protects its own systems. On the charter’s logic, aggregate patching performance across government and industry also shapes how much the government is willing to disclose. Faster patching across the ecosystem supports more disclosure, which in turn gives defenders more to patch.
Reading the questions as a defender’s scorecard
Each defensive question in Annex B has a counterpart inside a defending organization. Where a product is used and how widely maps to an asset inventory. How much users rely on it maps to a criticality rating. How soon a patch will be applied maps to time to patch. The share left unpatched for more than a year maps to a list of legacy and unpatchable systems. Whether exploitation can be detected maps to logging and monitoring coverage.
An organization that can answer these questions about itself is the kind of defender the charter hopes for. It knows where a vulnerable product runs, how much it matters, how fast a fix will land and what it will do where no fix can land. The charter does not ask defenders to keep such a scorecard. Its questions describe one, and in our reading a security program can borrow the structure directly.
The other equities in a defender’s terms
The commercial and international questions also have a defender’s reading. The commercial question asks how revealing government knowledge of a flaw would affect relationships with industry. For defenders, that relationship is what delivers patches and advisories. The international question concerns partner governments, whose own disclosures and advisories reach multinational banks and health systems. Both questions describe channels that defenders depend on.
The scorecard also helps with a question the charter leaves to defenders: which disclosed flaws to fix first. When several patches arrive at once, the same factors the government weighed, breadth of use, reliance, severity and exposure, can order the work. Using the same factors keeps an organization’s priorities aligned with the reasons the flaw was disclosed. It also makes the order of work explainable to auditors, because each choice traces to a stated factor rather than to habit.
Detection as a substitute
One mitigation question deserves attention from security operations teams. The charter asks whether exploitation can be detected by the government or other members of the defensive community. A restricted vulnerability whose use can be detected is less dangerous to leave unpatched. For defenders, this means detection capability is also part of the national calculation, alongside patching.
The clock: business days for decisions, a year for restrictions
The charter sets explicit timelines. They are short for deciding and disseminating, and longer for revisiting a decision to restrict.
Submission to determination
An agency submits a qualifying vulnerability as soon as practicable, with a recommendation. The secretariat notifies all points of contact within one business day of acknowledging the submission. Agencies with an equity must concur or not concur within five business days. Non-concurring agencies and the submitter hold discussions within seven business days to reach consensus. If they cannot, they give the board options.
The board aims for consensus. If it cannot reach consensus, it votes on a preliminary determination. An agency with an equity may contest that determination within five business days, and the dispute goes through the National Security Council process. Disclosure of a vulnerability preliminarily determined to be disclosable is delayed until the dispute is resolved.
Dissemination
Once disclosure is decided, the charter says dissemination should be as fast as possible and, when possible, within seven business days. The submitting agency is presumed to know the flaw best and is responsible for telling the vendor. It provides a copy of the dissemination to the secretariat.
Restriction and reassessment
A restricted vulnerability is reassessed annually by the board until it is disseminated, becomes publicly known or is otherwise mitigated. That is the longest clock in the charter. It means a flaw kept for operational use is revisited at least once a year. It also means it can remain restricted across several reassessments.
Annual reassessment from a defender’s side
The annual reassessment has a consequence for defenders that is easy to miss. A restricted flaw can remain unknown to its vendor across reassessments, and defenders cannot patch what the vendor does not know about. For that period, the only protections available are general ones: configuration baselines, segmentation, least privilege and detection. A program that relies on patches alone has nothing to offer against such a flaw.
The trigger for urgency
One rule overrides the ordinary clock. If an agency discovers ongoing malicious activity exploiting a vulnerability under a restriction decision, it must report it immediately. Discussion then begins no later than the next business day, and participants are to reach consensus quickly on disclosure or mitigation. Evidence that adversaries are using the flaw moves the decision forward.
That trigger connects the government’s process to defenders’ detection. If defenders and the government see exploitation in the wild, the charter’s clock speeds up. Incident reporting by agencies and their partners therefore feeds the process. In our reading, a mature incident reporting relationship with federal partners helps the government learn when a restricted flaw is being used against others.
Comparing the clocks with an agency’s own
The charter’s decision clock runs in business days. A defender’s internal clock for critical patches should be compared with it. If the government can decide and disseminate quickly, and the vendor can patch soon after, the remaining time to protection is the defender’s. An agency that needs months to install a critical fix adds months to every disclosure the government makes.
The comparison is inexact, because the charter sets no vendor clock and vendor timelines vary by product. Even so, it helps leadership. It shows that the slowest part of the chain from discovery to protection may lie inside the defender’s own estate. That part is where a security leader has direct control.
Clocks the charter does not set
The charter sets no clock for the vendor’s patch and none for installation. It expects the releasing agency to follow up so the board can judge whether the vendor’s action meets government requirements. If the vendor will not act, or not with urgency, the government may take other mitigation steps. The time from patch release to installation belongs to defenders, and the charter only asks the government to predict it.
The handoff: after disclosure, the work moves to vendors and defenders
The fourth part covers what disclosure sets in motion. The charter’s decision ends with information reaching a vendor. Protection depends on what happens next. Three parties act in sequence: the vendor builds a fix, the government follows up, and defenders install. The charter governs the middle step and predicts the other two.
The vendor’s part
The submitting agency tells the vendor and follows up. The charter asks the releasing agency to judge whether the vendor’s response meets government requirements. It anticipates vendors who choose not to address a flaw or who do not act with urgency. In those cases the government may take other steps, which may include engaging a broader community beyond government.
The defender’s part
Once a patch exists, the defender’s part begins. The patch reveals the flaw to anyone who studies it, which the charter’s reverse engineering question anticipates. From that moment, unpatched systems are exposed to a wider set of attackers. The charter’s calculation assumed that enough defenders would close that gap fast enough to make disclosure worthwhile.
WannaCry, in May 2017, was a public example of the gap. US-CERT’s alert said the attackers appeared to exploit a Windows SMB vulnerability for which Microsoft had released an update on March 14, 2017. The patch existed before the attack. In our reading, the systems at risk were those that had not installed it, and those running older Windows versions, for which the alert says Microsoft released patches on May 13.
Advisories as a signal
When a government agency itself publishes an advisory about a flaw it found, defenders learn two things. They learn about the flaw. They also learn, from the charter’s logic, that the government judged disclosure worthwhile, which implies an expectation that defenders will act quickly. An advisory from a government source is a cue to move the fix to the front of the queue.
The signal is strongest for flaws in widely deployed platforms, where the charter’s questions about breadth of use and patch uptake weigh heavily. For such flaws, the reverse engineering window opens as soon as the patch ships. Defenders who batch such a fix into a routine monthly cycle may give attackers the time the disclosure decision assumed they would not have.
What the handoff asks of agencies
For a federal civilian agency, the handoff becomes an obligation through other channels. In 2017, binding operational directives from DHS could require agencies to act. The charter itself did not set patch deadlines for agencies. Later directives did, as the hindsight section describes. The charter’s defensive logic and those later deadlines point the same way: the value of a disclosed flaw is realized in installation.
Components and downstream products
The charter’s scope includes open-source software and industrial control products. A flaw in a widely used component can affect products from several vendors. Disclosure to the component’s maintainer starts a chain: the maintainer fixes the component, product vendors adopt the fix, and defenders install the product updates. Each step adds time, and defenders stand at the end.
For defenders, this means a single disclosure can produce a series of patches over time, each for a different product. Knowing which products contain which components becomes part of patching. The charter does not address this directly. Its question about how broad a range of products or versions is affected shows that component spread is one of the facts weighed.
Mitigation without disclosure
The charter’s option to share mitigation information without disclosing the vulnerability also lands on defenders. An agency or company might receive guidance to change a configuration or block a protocol without being told why. Acting on such guidance quickly, without the full story, is part of the handoff. Defenders who wait for a patch or a full explanation may miss the protection the government chose to offer.
The handoff for organizations that receive reports
The charter also says the government will safeguard vulnerability information it receives from businesses, researchers and foreign governments, and encourage them to disclose consistent with international standards. That puts agencies and companies on the receiving end of reports too. An organization that runs its own vulnerability disclosure intake is part of the same chain, as a vendor of its own software and services.
How this reached different organizations
The charter binds federal agencies. Its logic reaches others through the patches and guidance it produces. We take six kinds of organization in turn.
The charter’s reach beyond government runs through three channels. Patches reach every user of a product. Mitigation guidance can reach chosen entities without a patch. And the assumptions the government makes about defenders shape what it chooses to disclose. Each kind of organization below meets those channels differently.
Federal civilian agencies
Civilian agencies participate in the process if they hold equities, and they are defenders of their own systems. Commerce, Treasury, State, Justice, Homeland Security and Energy sit on the board. For a civilian agency security leader, the practical points are internal. Know who the agency’s VEP point of contact is, and route vulnerability findings from staff or contractors to that person. Treat patch performance as part of the agency’s contribution to the national calculation.
Defense and intelligence agencies
Defense and intelligence agencies hold both offensive and defensive equities. The charter makes NSA the secretariat and requires flaws in NSA-certified government products or cryptographic functions to be reported to NSA. These agencies’ internal processes for deciding what to submit are their own. Congress later required the intelligence community to report on them, as the hindsight section shows.
State and local governments
State and local governments stand outside the process, as defenders of systems that run the same commercial products. When a disclosed flaw is patched, their patch speed is part of the uptake the charter predicts. A county with a small security team faces the same reverse engineering window as a federal agency. Federal patch directives did not bind them.
Banks and credit unions
Financial institutions are part of the business population the charter’s patch question names. Their patching is governed by their own regulators’ expectations and their risk programs. A bank that patches critical systems quickly, and can show it, contributes to the uptake the charter assumes. Treasury sits on the board, though how any department argues its views is not public.
For a bank, the useful habit is to treat government advisories and vendor patches for widely used platforms as one category with a short internal deadline. A credit union with a small team may rely on its core processor and managed service providers for much of its patching. In that case, the contract with each provider is where patch speed is set, and the institution should know what those contracts promise.
Health systems
Health systems run commercial software alongside medical devices and clinical systems that can be hard to patch. The charter’s question about systems that remain unpatched for more than a year describes part of a hospital’s estate in our reading. For those systems, the mitigation questions matter more: configuration changes, network segmentation and detection may be the only protection available when a disclosed flaw cannot be patched promptly.
This gives hospital security teams a reason to keep a current list of clinical systems that cannot take vendor patches on the normal cycle, with the compensating control for each. When a disclosure lands for a platform those systems use, the team can see at once which ones depend on mitigation rather than a fix. The charter’s question about long-unpatched systems describes exactly the population that list covers.
Vendors that sell to government
Technology vendors are the direct recipients of disclosures. The charter expects a vendor that receives a report to act with urgency consistent with the risk, and the releasing agency follows up. For vendors that sell to agencies, banks and hospitals, the charter is a reason to keep a clear intake for government reports and a documented response process. The government’s follow-up may ask what the vendor did.
Objections, in order
The argument that the charter made patch uptake a factor in disclosure faces objections. We take the strongest six in order.
1. Patch uptake is one question among several and carries little weight
The charter assigns no weights, so we cannot say how much the question counts. Our argument is that it is there, in writing, as a defensive consideration. Its presence tells defenders that their behavior is modeled. That holds whatever weight it carries in a given case.
2. The charter defaults to disclosure, so patch speed seldom comes into play
The charter does say disclosure is in the national interest in the vast majority of cases. If so, the patch questions shape a minority of decisions. They still shape those, and those are the hardest cases. Defenders gain from fast patching in every case, and in the hard cases their speed is part of the argument.
3. Individual defenders cannot affect a national estimate
One agency or bank cannot move a national estimate on its own. Aggregate behavior is made of individual behavior, though, and federal directives later set shared deadlines across civilian agencies. A security leader can also affect the part of the estimate that covers their own sector by sharing patch performance through sector channels.
4. The process is secret, so its logic tells defenders nothing
The individual decisions are secret. The considerations are public, and that was the point of the release, according to the White House article. A public framework lets defenders understand what the government weighs, even when they cannot see a single case. That understanding is enough to inform how defenders think about their own patch speed.
5. Vendors, not defenders, are the bottleneck
Vendor speed matters, and the charter addresses it through follow-up and possible further steps. But the WannaCry example shows a case where the vendor’s patch predated the attack. In our reading, the bottleneck that day was installation. The charter’s questions cover both, and defenders control only one of them.
6. Patch speed is already a priority, so the charter adds nothing
We think patch speed was already a priority for most security programs before 2017. What the charter adds is a reason that reaches beyond the organization. It shows that the government’s own disclosure decisions take defenders’ behavior into account. For a leader seeking budget for patching, that is a different argument from local risk reduction, and in our reading a stronger one.
What an agency security leader could check
The charter suggests a handful of checks for a federal agency, and they transfer to state agencies, banks and hospitals:
- Name the agency’s VEP point of contact, and make sure staff and contractors know to route new vulnerability findings there.
- Write contract language that sends vulnerability findings from testing and research contractors to the agency, not to the contractor’s discretion.
- Measure the time from release to installation for critical patches, by system class, and report it to leadership.
- Keep a list of systems that cannot be patched promptly, with the compensating configuration, segmentation and detection for each.
- Set a fast path for acting on government mitigation guidance that arrives without a full explanation.
Reporting is the part these checks share. Consider a security leader who can show time to patch by system class, the unpatchable systems with their compensating controls, and the path for mitigation guidance. That leader has the evidence the charter’s questions imply. That evidence also answers the questions an inspector general or auditor would ask about vulnerability management.
Each check strengthens the agency’s own defense. Taken together across agencies and companies, they also strengthen the assumption the charter makes when it weighs disclosure. In our view, that shared effect deserves more attention than it receives. The charter treats defenders’ patch speed as part of national security policy, alongside its role as hygiene.
With hindsight
On December 19, 2017, the White House publicly attributed the WannaCry attack to North Korea. The press briefing said the malware had encrypted and disabled computers in hospitals, schools, businesses and homes, and had affected the UK’s health care system. A DHS official at the briefing said WannaCry began on May 12, 2017. The attribution gave the patching gap of that spring a named adversary.
Congress then named the charter, by title and date, in statute. The National Defense Authorization Act for Fiscal Year 2020 was approved on December 20, 2019. Its section 6720 defined the Vulnerabilities Equities Policy and Process document as the executive branch document of that title dated November 15, 2017. It defined the process as the interagency review under that document or any successor document. It required the Director of National Intelligence to report to the intelligence committees, within 90 days, on how each intelligence community element decides what to submit. Each element must report any significant change to that process within 30 days.
The same section required annual classified reports to the intelligence committees. They must give the number of vulnerabilities submitted, the number disclosed to vendors or the public, and the aggregate number excluded under the charter’s exceptions paragraph. Each report must include an unclassified appendix with the aggregate numbers disclosed and the numbers disclosed that are known to have been patched. Congress thus asked for patch outcomes to be counted.
That last requirement fits our argument. In our reading, the statute treats the patch, and not the disclosure alone, as an outcome worth counting. It asks for the number of disclosed flaws known to have been patched. The reports go to committees, not the public, and we do not describe their contents.
One public disclosure showed the handoff at work. On January 14, 2020, NSA published an advisory saying it had discovered a critical vulnerability, CVE-2020-0601, in Windows cryptographic functionality. The advisory said rapid adoption of the patch was the only known mitigation and should be the primary focus for all network owners. Where enterprise-wide automated patching was not possible, it said owners should prioritize endpoints that provide essential or broadly relied-upon services. The advisory did not describe the internal process behind the disclosure, and we do not infer it.
The defender side of the handoff became a federal requirement. On November 3, 2021, CISA issued Binding Operational Directive 22-01. It created a catalog of known exploited vulnerabilities and required federal civilian agencies to remediate listed flaws on set timelines. The default was two weeks for vulnerabilities with identifiers assigned from 2021, and six months for older ones. The directive turned installation speed for exploited flaws into a measured duty.
On June 10, 2026, CISA issued Binding Operational Directive 26-04, which revoked BOD 22-01 and an earlier directive and replaced them. The new directive sets remediation urgency by four variables. They are whether the asset is publicly exposed, whether the flaw is in the known exploited catalog, whether an adversary can automate the exploit, and whether exploitation gives partial or total control. Its background says adversaries’ use of AI may further narrow the time between patch release and exploitation.
In our reading, BOD 26-04 reads, in part, like an answer to the charter’s defensive questions from the defender’s side. The charter asked how widely a product is used, how severe the flaw is, and how fast patches will be installed. The directive asks agencies to decide patch speed by exposure, exploitation and impact. Both documents treat time to installation as the variable that turns a disclosed flaw into protection.
What held up, then, is the link between disclosure and installation. The charter wrote it into the government’s decision. Congress asked for patch outcomes to be counted. CISA made installation speed a binding duty for civilian agencies and then tied it to risk. What we cannot assess is how the equities process itself performed, because its case decisions remain classified. Defenders in 2026 can still act on the part of the framework they control.
Spotted an error? Tell us and we will check it and correct the article where needed.