EO 13800 asked agency heads to record the cyber risks they accepted and the budget choices behind them

On May 11, 2017, an executive order made agency heads accountable for cyber risk and asked each to document accepted risk with its budgetary reasons. In our reading, that turned risk acceptance into a spending record OMB could compare across agencies.

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 May 11, 2017, an executive order made agency heads accountable for cyber risk and asked each to document accepted risk with its budgetary reasons. In our reading, that turned risk acceptance into a spending record OMB could compare across agencies.

  1. The week Section 1 opened with a policy statement.
  2. The signature: who could accept risk for an agency The first part asks who decides.
  3. The ledger: what the record had to contain The second part asks what the agency wrote down.
  4. The aggregate: from agency records to a government-wide view The third part asks what OMB and DHS did with the records.
  5. The purchase: shared services as the default The fourth part asks how the order steered buying.
  6. How the order reached different kinds of agency The four parts applied across the executive branch.
  7. Objections, in order It did, and the order said it was acting in accordance with that law. [1] [3] M-17-25 described the order as reinforcing FISMA. [3] The new element was the specific record of accepted risk with its budgetary reasons, reviewed in aggregate. [1] Responsibility existed before.
  8. What agency budget and technology leaders could check The four parts translate into checks any public body can run on its own cyber risk decisions, federal or not.
  9. With hindsight OMB published the Federal Cybersecurity Risk Determination Report and Action Plan in May 2018. [4] It said it was issued in accordance with Executive Order 13800 and M-17-25. [4] Its results are set out in the body above.

The week

On May 11, 2017, the President issued an executive order titled Strengthening the Cybersecurity of Federal Networks and Critical Infrastructure. [1] It had three operative sections: the cybersecurity of federal networks, the cybersecurity of critical infrastructure, and cybersecurity for the nation. [1]

Section 1 opened with a policy statement. The executive branch operates its information technology on behalf of the American people, and the President will hold agency heads accountable for managing cybersecurity risk to their enterprises. [1] Because one agency’s choices can affect risk to the whole executive branch, it was also policy to manage cybersecurity risk as an executive branch enterprise. [1]

The order then set out findings. The executive branch had for too long accepted antiquated and difficult-to-defend IT. [1] Known but unmitigated vulnerabilities were among the highest cybersecurity risks agencies faced. [1] The order defined them to include using operating systems or hardware beyond the vendor’s support lifecycle, declining to implement a vendor’s security patch, and failing to follow security configuration guidance. [1] Effective risk management, it said, required agency heads to lead integrated teams with expertise in IT, security, budgeting, acquisition, law, privacy and human resources. [1]

The risk management subsection held the core duties. Agency heads would be held accountable for risk management measures commensurate with the risk and magnitude of harm from unauthorized access, use, disclosure, disruption, modification or destruction of IT and data. [1] They would also be held accountable for aligning cybersecurity risk management with strategic, operational and budgetary planning. [1]

Effective immediately, each agency head was to use the NIST Framework for Improving Critical Infrastructure Cybersecurity, or any successor document, to manage the agency’s cybersecurity risk. [1] Within 90 days, each agency head was to send a risk management report to the Secretary of Homeland Security and the Director of OMB. [1]

The report had two required elements. It had to document the agency head’s risk mitigation and acceptance choices, including the strategic, operational and budgetary considerations behind them and any accepted risk, including from unmitigated vulnerabilities. [1] It also had to describe the agency’s action plan to implement the Framework. [1]

DHS and OMB would then jointly assess the reports and determine whether the choices were appropriate and sufficient to manage risk to the executive branch in the aggregate. [1] Within 60 days of receiving the reports, OMB was to send the President that determination and a plan. [1] The plan was to protect the enterprise, address immediate unmet budgetary needs, set a regular process for reassessment and recurring budget needs, and reconcile and align federal security policies with the Framework. [1]

The order also set a direction for architecture. Agency heads were to show preference in procurement for shared IT services, including email, cloud and cybersecurity services. [1] The Director of the American Technology Council was to coordinate a report on federal IT modernization within 90 days, covering consolidated network architectures and shared services. [1] For national security systems, the Secretary of Defense and the Director of National Intelligence would implement the order instead of DHS and OMB, with a report within 150 days. [1]

Sections 2 and 3 called for a series of reports. They covered support to critical infrastructure at greatest risk, market transparency of cyber risk management by critical infrastructure entities, botnets, electricity disruption, the defense industrial base, deterrence, international cooperation and the cybersecurity workforce. [1] The deadlines ranged from 45 days for the international reports to one year for the final botnet report. [1]

Federal agencies already had a security law. OMB later noted that both the Federal Information Security Modernization Act of 2014 and the order identify the agency head as the official responsible for each agency’s cybersecurity in the end. [4] The order cited that law’s chapter of title 44. [1] What the order added was a specific record. Each agency head had to write down which risks the agency was mitigating, which it was accepting, and the budgetary considerations behind those choices. [1] The record then went to OMB and to DHS. [1]

Our argument is that this made risk acceptance a budget document. An accepted risk is a decision not to spend, or to spend elsewhere. The order asked agencies to say so in writing, and asked OMB to compare those choices across the government and to plan for unmet budgetary needs. [1] The test of the order, in our reading, was whether that record would change where money went.

We take the order’s federal network provisions in four parts:

  1. The signature: who in an agency could accept risk on its behalf.
  2. The ledger: what the agency had to record about its choices.
  3. The aggregate: how OMB and DHS turned agency records into a government-wide determination.
  4. The purchase: how the order steered spending toward shared services.

The first part is about authority. The second is about the content of the record. The third is about review. The fourth is about buying. Together they run from the person who decides to the contracts that carry the decision out. After the four parts, we look at how the order reached different kinds of agency, answer objections in order, and list what agency budget and technology leaders could check.

A note on sources and our reading. We describe the order from the White House text of May 11, 2017, and the Federal Register text published May 16, 2017. [1] [2] We describe OMB’s implementing guidance from M-17-25, and its results from OMB’s May 2018 risk report. [3] [4] Statements about how the order affected particular agencies’ budgets are our reading, not findings in those documents.

The signature: who could accept risk for an agency

The first part asks who decides. The order’s answer was the agency head, held accountable by the President. [1] OMB’s guidance said how that worked in practice.

The senior accountable official

M-17-25 required each agency to notify OMB, on or before May 26, 2017, of the senior accountable official responsible for implementing section 1(c) of the order. [3] That official could be the agency head or a designee. [3] A designee had to meet four conditions. [3] The person had to report directly to the agency head and have vision into all areas of the organization. [3] The person also needed authority over both funding and management of IT and enterprise risk, and the ability to represent challenges across the enterprise. [3]

The third condition is the one that mattered for budgets. M-17-25 tied the official’s role to aligning information security management with strategic, operational and budgetary planning. [3] An official without budget authority could describe risk. Only an official with budget authority could decide to accept it or pay to reduce it.

Why the CIO was not enough

OMB’s later report described a pattern of delegation. It said agency heads often delegated cyber risk management to the CIO and CISO, and many agencies could not describe leadership engagement above the CIO level. [4] The report also said CIOs and CISOs often lacked the authority to make organization-wide decisions. [4]

The order’s design anticipated that gap, in our reading. By naming the agency head and requiring a designee with funding authority, it put the signature where the money was controlled. [1] [3] Whether each designee met the test in substance, as well as on paper, is a separate question. OMB’s 2018 findings suggest some did not. [4]

The written response

M-17-25 added a second signature. OMB would give agencies an initial assessment by July 28, 2017. [3] Each agency then had to provide a written response, signed by the agency head or the senior accountable official, describing how it planned to accept, avoid, transfer or mitigate outstanding risks, by August 9, 2017. [3]

That response is the heart of the matter. It forced a choice among four options for each outstanding risk. Acceptance is a legitimate choice. The guidance required that it be made by a named official and written down. [3]

Four options, four budget meanings

The four options in M-17-25 each have a budget meaning, in our reading. [3] To mitigate is to spend on a control. To avoid is to stop the activity that creates the risk, which may save money or cost a mission. To transfer is to move the risk to another party, such as a shared provider or an insurer, through a contract with its own cost. To accept is to spend nothing now and carry the exposure.

Laid out that way, the written response was a budget choice for each outstanding risk. An agency that answered mitigate for everything was implicitly asking for money. An agency that answered accept for everything was implicitly saying the budget would not change. Neither answer was credible across the board. The response forced the agency to sort its risks by what it would pay to change them.

What the signature changed inside an agency

A signed acceptance changes internal incentives. A CISO who reports a gap can now ask whether the agency head will sign for it. A budget office that declines a security request can be asked to record that the risk is accepted. In our reading, the signature gave security staff a way to move a disagreement up the chain without escalating it as a complaint. The decision went to the person the order held accountable. [1]

When the signer changes

Agency heads change with administrations, and designees can change at any time. In our reading, a signed acceptance is only as durable as the process that renews it. An acceptance signed by one official and never revisited becomes an inherited risk that no current official has chosen. M-17-25 required the risk assessments themselves to continue on a semiannual or annual cycle. [3] The memo did not say the signed response had to be renewed; in our reading, each cycle was still the natural point to renew it, and a chance for a new official to take ownership of the old choices, or to change them.

The ledger: what the record had to contain

The second part asks what the agency wrote down. The order required documentation of mitigation and acceptance choices, the considerations behind them, accepted risk including from unmitigated vulnerabilities, and an action plan for the Framework. [1] OMB turned those requirements into a reporting format.

FISMA metrics as the risk report

M-17-25 did not create a new questionnaire. All agencies, including small ones, had to submit responses to the fiscal year 2017 third-quarter FISMA CIO metrics through DHS’s CyberScope system on or before July 14, 2017. [3] Those submissions served as the risk management report the order required. [3] OMB said the choice was made to minimize reporting burden. [3]

That choice had consequences. The FISMA CIO metrics were already aligned with the Framework’s function areas, according to the memo. [3] So the risk report described capabilities against a common structure. It did not ask each agency to invent its own description of risk. The aggregate review in the third part depended on that common structure.

The Framework action plan

The second element was an action plan to implement the Framework, due through CyberScope by July 14, 2017. [3] M-17-25 listed what it had to include. One item was the status of IT budget materials aligned with the Framework under the fiscal year 2018 IT Budget Capital Planning Guidance. [3] Others were proposed internal use of updated metrics, a timeline to map capabilities to the Framework functions, and proposed use of the Framework’s terminology. [3]

The memo said OMB had aligned the IT security portion of the fiscal year 2018 capital planning guidance with the Framework. [3] It said the alignment helped standardize vocabulary and definitions used in security. [3] In our reading, this was the mechanism that made the ledger a budget ledger. A spending request and a risk acceptance could now be described in the same terms, against the same five functions: Identify, Protect, Detect, Respond and Recover. [3]

Unmitigated vulnerabilities as an entry

The order singled out one kind of accepted risk: unmitigated vulnerabilities. [1] Its findings listed what counted. Unsupported operating systems or hardware, unapplied vendor patches and unfollowed configuration guidance were all known vulnerabilities. [1] Each is a choice with a cost on both sides. Replacing unsupported systems costs money. Keeping them carries risk.

The order put that trade in the ledger. An agency that kept an unsupported system for a mission reason now had to say so as an accepted risk, with its budgetary reason. [1] That entry is a statement about money as much as about security.

What the ledger did not contain

The ledger had limits. The FISMA metrics described capabilities, not threats. OMB’s 2018 report found agencies were not equipped to determine how threat actors sought access to their information. [4] It said the risk-based budgeting model then in use treated missing capabilities as perceived risk, without assessing the likelihood of those risks materializing. [4] A ledger built on capabilities could show what an agency lacked. It could not show which gap an adversary was most likely to use.

How an agency could build the ledger

The order gave agencies a short window. [1] In our reading, an agency that already kept a current system inventory, plans of action for known weaknesses and a link between systems and budget lines could assemble the ledger from existing records. An agency without those records had to build them first.

The practical sequence was the same either way. List the systems and the capabilities the FISMA metrics asked about. [3] Mark the gaps. For each gap, record whether a fix was funded, planned or not planned. For each unplanned fix, record why: cost, mission need, vendor support, or something else. The last column was the new one. It turned a list of gaps into a list of decisions.

The comparison with public companies

OMB’s 2018 report made a comparison of its own. It noted that the Securities and Exchange Commission requires publicly traded companies to file quarterly and annual reports to inform shareholders of risks, including cyber risks. [4] It said federal agencies would benefit from a similar process that tracks quarterly performance against strategic targets, communicates the resulting risks and gives a sense of return on investment for cybersecurity protections. [4] In our reading, that is a description of the ledger the order started, made recurring.

What the ledger meant for an agency’s finance staff

For a chief financial officer, the order brought security into budget formulation in a recordable way. A budget justification for a security investment could cite a gap in the Framework-aligned metrics. A decision not to fund it would leave a corresponding accepted risk in the record. In our reading, the two documents could now be read side by side, at least in principle. The practical question was whether budget offices and security offices used the same identifiers for the same gaps.

The aggregate: from agency records to a government-wide view

The third part asks what OMB and DHS did with the records. The order required them to determine whether agency choices were appropriate and sufficient for the executive branch in aggregate, and to plan for unmet budgetary needs. [1]

The enterprise idea

The order’s policy said one agency’s risk choices can affect the executive branch as a whole and national security. [1] That is the reason for an aggregate view. A small agency that accepts a risk on a shared network connection may pass it to others. The order made the executive branch, not each agency, the unit of risk management for that purpose. [1]

The methodology

M-17-25 said OMB and DHS would use the FISMA metrics to produce agency-specific risk assessments and include them in the report to the President. [3] Chief Financial Officer Act agencies would continue such assessments semiannually and small agencies annually, to track performance against the Framework. [3] OMB’s 2018 report described the result. OMB and DHS assessed 96 agencies across 76 metrics, including four narrative responses. [4] They rated each agency’s program as managing risk, at risk or high risk. [4]

The determination

The determination was stark. OMB and DHS found 71 of 96 agencies, 74 percent, had programs that were either at risk or high risk. [4] The table split the 96 as 25 managing risk, 59 at risk and 12 high risk. [4] The report also found agencies were not equipped to determine how threat actors seek access, and that the lack of threat information resulted in ineffective allocation of limited cyber resources. [4]

What the aggregate said about money

The report’s budget findings give direct support to our argument. Civilian agencies projected fiscal year 2017 spending of $5.7 billion on cyber defenses across the Framework functions, against $5.0 billion in fiscal year 2016. [4] OMB said the increase came without a sense of prioritization or return on investment in reducing risk. [4]

It gave an example. Fewer than 16 percent of agencies met the target for encrypting data at rest. [4] Non-defense agencies budgeted less than $51 million on encrypting data at rest in fiscal year 2017, with half of that coming from two agencies. [4] By comparison, agencies had budgeted almost $210 million for attaining and renewing authorities to operate. [4] OMB concluded that the government’s priorities must be realigned. [4]

In our reading, those numbers are what the ledger made visible. Without a Framework-aligned budget vocabulary, OMB could not have compared encryption spending to authorization spending across agencies. [3] [4] With it, a pattern of accepted risk showed up as a pattern of spending.

The risk-based budgeting model

The report described the tool OMB built to connect the ledger to money. OMB said it had spent the past two years developing a risk-based budgeting model that ties agency cybersecurity spending to the FISMA metrics. [4] It planned to disseminate the process in early fiscal year 2018 so agency CIOs, CISOs and CFOs could communicate cyber risks and budget for the most critical needs. [4]

OMB was candid about the model’s limit. In its then-current form, it treated missing capabilities as perceived risk and did not assess the likelihood of those risks materializing. [4] OMB said adding a cyber threat framework would let an agency identify strong and weak coverage against known threats and give a clearer return on investment. [4] That is the ledger growing a threat column.

Other gaps the aggregate exposed

The aggregate review found gaps that each carried a budget meaning. Inspectors general reported that 55 percent of agencies limited access based on user attributes and roles, and 57 percent reviewed and tracked administrative privileges. [4] Inspectors general found 49 percent of agencies could detect and whitelist software running on their systems. [4] Only 27 percent of agencies reported the ability to detect and investigate attempts to access large volumes of data. [4] In our reading, each figure described a capability the government was, in effect, choosing not to fund at scale.

The shared program behind the aggregate

DHS’s Continuous Diagnostics and Mitigation program was part of the same picture. The report said Phase 1, which DHS initiated in 2015, focused on hardware, software, vulnerability and secure configuration management. [4] At that stage, 62 percent of agencies had capabilities in place through the program to understand the devices accessing their networks. [4] Phase 2, originally planned for full deployment by December 2016, was then being deployed. [4]

The program matters for our argument because it is a shared service in the security sense. A capability bought once and deployed to many agencies is the purchase part in action. Its delays also show the trade-off. When a shared program slips, every agency depending on it carries the gap, and each agency’s ledger has to record a risk whose fix sits with someone else. [4]

The four actions

The report set four actions. Implement a cyber threat framework to prioritize efforts. Standardize IT and cybersecurity capabilities to control costs and improve asset management. Consolidate agency security operations centers. And drive accountability through governance processes, recurring risk assessments and engagement with agency leadership. [4] The second and third actions are procurement actions as much as security ones. They lead into the fourth part.

The purchase: shared services as the default

The fourth part asks how the order steered buying. Agency heads were to show preference in procurement for shared IT services, to the extent permitted by law, including email, cloud and cybersecurity services. [1]

A preference, not a mandate

The order’s wording was a preference. [1] It did not require any agency to move any service. It did change the default question in an acquisition. A program office proposing its own email system, or its own security monitoring, now had to explain why a shared service would not do. In our reading, that shifted the burden of justification in budget reviews.

The modernization report

The order also called for a report on transitioning agencies to consolidated network architectures and shared services, including email, cloud and cybersecurity. [1] The report was to address legal, policy and budgetary considerations, technical feasibility, cost effectiveness, timelines and milestones. [1] All agency heads had to supply information about their current IT architectures and plans. [1] That information request was itself an inventory exercise, and a budgetary one.

Standardization and consolidation as cost controls

OMB’s 2018 report connected the threads. It found a lack of standardized IT capabilities and limited network visibility. [4] It said agencies continued to buy single point solutions for perceived gaps rather than funding gaps that threat actors were actually exploiting. [4] Its planned actions included standardizing capabilities to control costs and consolidating security operations centers. [4] In our reading, those are the shared services preference expressed as a spending plan.

Email as the worked case

The report gave email as an example of the purchase part at work. The 23 civilian Chief Financial Officer Act agencies had nearly 2.2 million email inboxes between them. [4] One agency listed no fewer than 62 separately managed email services. [4] OMB said that made it virtually impossible to track and inspect inbound and outbound communications across that agency. [4]

At least nine of the 23 agencies had already consolidated enterprise email. [4] Agencies that had consolidated, or were consolidating, identified savings in the range of $1 million to $4 million per year. [4] OMB said it would work with agencies on enterprise-wide email consolidation plans. [4] In our reading, email showed the purchase part and the ledger part meeting. Consolidation reduced cost and also made a security capability, inspection of messages, possible.

Security operations centers

The report also took up security operations centers. It said some agencies lacked enough skilled staff to operate one effectively, and others ran multiple centers with different processes and technology. [4] Its action was for agencies to begin consolidating their centers, or migrating to a security operations center as a service, in fiscal year 2019. [4] That is a procurement decision with a security rationale. It is also the shared services preference applied to a labor-intensive part of defense.

What procurement offices had to change

For an agency’s acquisition staff, the preference meant new homework. Market research had to include the shared services already available to the agency. Business cases had to compare an agency-specific build with a shared offering on cost and security. Contracts for agency-specific tools needed a stated reason. In our reading, this also raised the bar for evidence about shared providers. If an agency relied on a shared security service, its own risk report had to describe that dependence, because the accepted risk now partly sat with someone else.

The cost of moving

A preference for shared services has a transition cost that a budget has to carry. In our reading, it has four pieces. The agency runs old and new services in parallel for a period. It migrates data and users, which takes staff time. It renegotiates or ends existing contracts, which may carry termination terms. And it retrains staff on the new service. These costs do not appear in the steady-state savings figures that make consolidation attractive.

An honest ledger would record them. If an agency deferred a consolidation because it could not fund the transition, the risk it kept was an accepted risk with a budgetary reason, which is exactly what the order asked agencies to write down. [1]

The trade-off

Shared services carry their own risk. Concentration means one provider’s failure affects many agencies. In our reading, the order accepted that trade in exchange for consistency and visibility. The aggregate view in the third part is easier to build when many agencies use the same tools. The ledger in the second part is easier to compare when the capabilities are the same.

How the order reached different kinds of agency

The four parts applied across the executive branch. They played out differently by agency type. The descriptions below are our reading of how the structure fit each setting.

Large departments with component CIOs

OMB’s report singled out federated agencies, which employ multiple component CIOs who often control their own budgets. [4] For a department of that shape, the signature part was the hardest. A department head could sign a risk report while the budget that created the risk sat in a component. The order’s requirement for a designee with funding authority pointed at that problem. [3] It did not restructure the department’s budget lines.

Small agencies

Small agencies had the same reporting duties, including the FISMA metrics and the Framework action plan. [3] They would assess annually rather than semiannually. [3] With small staffs, the ledger was a heavier lift relative to their size. Shared services offered them capabilities they could not build alone.

National security systems

For national security systems, the Secretary of Defense and the Director of National Intelligence implemented the order instead of DHS and OMB, and had to justify any deviation. [1] Their reports could be classified. [1] The four parts still applied. The audience for the ledger was different.

Inspectors general

OMB’s methodology drew on inspector general FISMA metrics as well as CIO metrics. [4] An agency’s accepted risks therefore met a second, independent description of its program. Where the two disagreed, the gap itself was information for OMB and for agency leadership. For an agency’s own staff, the inspector general’s view was a preview of how the ledger would be read. An accepted risk that the inspector general also reported as a weakness needed a stronger written reason than one the inspector general had not flagged.

Contractors and service providers

The order applied to agencies, not to their contractors. Its effects reached contractors through the purchase part. A contractor running a system on an agency’s behalf held part of the agency’s accepted risk. An agency preferring shared services would buy fewer bespoke systems and more standard ones. In our reading, vendors that could show Framework-aligned evidence were better placed to answer the agencies’ new questions.

Agencies with old systems

Agencies running unsupported systems faced the order’s findings directly. The order named systems beyond the vendor’s support lifecycle as a known vulnerability. [1] An agency keeping such a system for a mission reason had to carry it in the ledger as accepted risk. In our reading, that entry made the replacement cost visible every time the ledger was reviewed. A modernization request could point to it. A decision to defer replacement had to be renewed, not left to lapse into habit.

State and local governments

The federal network provisions did not bind states or localities. The critical infrastructure section called for reports on support to critical infrastructure entities and consultation with state, local, tribal and territorial governments on electricity disruption. [1] For state budget offices, the federal ledger offered a model rather than a mandate: name the official who accepts risk, record the reason, and link it to the budget.

Objections, in order

1. FISMA already made agency heads responsible

It did, and the order said it was acting in accordance with that law. [1] [3] M-17-25 described the order as reinforcing FISMA. [3] The new element was the specific record of accepted risk with its budgetary reasons, reviewed in aggregate. [1] Responsibility existed before. A comparable written trail of accepted risk did not.

2. Using FISMA metrics made the report a compliance exercise

The metrics did measure capabilities, not threats, and OMB later said so. [4] They also made comparison possible, which is what produced the spending findings. [4] The fair reading is that the first round of the ledger was a capabilities ledger. OMB’s planned threat framework was meant to add the missing side. [4]

3. Accepting risk in writing invites blame

That may discourage candor. It also prevents silent acceptance, which is worse. The order’s design assumed agency heads would rather record a reasoned acceptance than be held accountable for an unrecorded one. [1] Whether candor held up depended on how OMB used the records. The report described quarterly assessments and engagement with agency leadership. [4]

4. The order did not come with money

It did not appropriate funds. It was to be implemented subject to the availability of appropriations. [1] What it did was require a plan to address unmet budgetary needs. [1] That plan rested on the ledger. In our reading, an order cannot appropriate, so the best it can do is make the case for money visible.

5. Shared services concentrate risk

They can. The order accepted that trade, and the modernization report was meant to weigh it, including effects on cybersecurity. [1] The alternative, many agency-specific tools, was the pattern OMB later criticized as single point solutions for perceived gaps. [4]

6. The Framework was voluntary guidance for critical infrastructure, not federal agencies

The Framework’s title names critical infrastructure. [5] The order made its use mandatory for agencies, effective immediately. [1] It also allowed for successor documents, which mattered as the Framework changed. [1] [5] [7]

7. Seventy-four percent at risk shows the order failed

The 2018 figure described the state of programs when the order’s process measured them. [4] It did not measure the order’s effect, because the order’s first product was the measurement itself. A baseline that shows most programs at risk is uncomfortable. It is also what an honest ledger would show at the start, in our reading.

8. The ledger was paperwork that changed nothing

The record’s value depends on use. OMB said the actions would help shape agency budgets for fiscal year 2019 and beyond, and it planned quarterly assessments reported to deputy secretaries. [4] Those are uses tied to budget and leadership attention. Whether they changed specific funding decisions at a given agency is beyond what the public record shows, and we do not claim it.

What agency budget and technology leaders could check

The four parts translate into checks any public body can run on its own cyber risk decisions, federal or not.

  1. Is there a named official, with authority over both funding and IT, who signs for accepted cyber risk?
  2. Does each accepted risk record the budgetary reason it was accepted and the cost of mitigating it?
  3. Do budget requests and risk records use the same identifiers, so a reader can match a funding decision to a risk?
  4. Can the organization compare its security spending by function with a peer’s or a central baseline?
  5. When a shared service is chosen, does the risk record say which risks now depend on the provider?

The checks work at any scale. A county with one IT director can name the signer, write the reason beside each accepted risk, and match it to the budget line. A large department needs the same three things, multiplied by its components. In both cases, the record has to be renewed when the budget is, or it goes stale.

The second check is the one the order made distinctive. A list of gaps is common. A list of gaps with the budgetary reason each was accepted is a document a budget office has to answer. [1]

With hindsight

OMB published the Federal Cybersecurity Risk Determination Report and Action Plan in May 2018. [4] It said it was issued in accordance with Executive Order 13800 and M-17-25. [4] Its results are set out in the body above. OMB and DHS found 71 of 96 agencies at risk or high risk, and set four actions on threat awareness, standardization, consolidation of security operations centers and accountability. [4]

The report described how the signature part had worked. At the end of the process, OMB required each agency’s senior accountable official to submit a signed letter describing the agency’s plan to accept, mitigate, avoid or transfer cybersecurity risks in the near term. [4] It also said OMB used information from its management and budget oversight processes to contextualize the findings. [4]

On accountability, OMB planned to require Chief Financial Officer Act agencies to submit risk assessments quarterly, with results provided to deputy secretaries through the President’s Management Council. [4] Small agencies would participate semiannually. [4] OMB said it would use its improved visibility into agency cybersecurity spending to make sure agencies invest in priority protections. [4] It said the actions would help shape agency budgets for fiscal year 2019 and beyond. [4]

NIST published version 1.1 of the Framework on April 16, 2018. [5] Under the order’s wording, a successor document would govern agencies’ use. [1] In our reading, that clause meant the order’s mandate moved with the Framework without a new order.

On May 12, 2021, Executive Order 14028, Improving the Nation’s Cybersecurity, set a new federal direction. [6] Its section on modernizing federal cybersecurity called for agencies to advance toward Zero Trust Architecture, accelerate movement to secure cloud services, centralize access to cybersecurity data, and invest in technology and personnel. [6] Within 60 days, each agency head was to update plans to prioritize resources for cloud adoption and develop a Zero Trust implementation plan with a schedule. [6] The 2021 order again placed duties on the agency head and tied them to resource plans. [6] It added specific technical destinations that the 2017 order had left open.

NIST published version 2.0 of the Framework on February 26, 2024. [7] It describes itself as a taxonomy of high-level cybersecurity outcomes for organizations of any size or sector, and lists cybersecurity risk governance among its keywords. [7] For agencies still bound by the 2017 order’s Framework requirement, the successor clause pointed to this version. [1] [7] That last point is our reading, and agencies should confirm current OMB direction.

What held up from May 2017? The record did. OMB’s 2018 report could compare encryption spending with authorization spending across agencies because the ledger had a common vocabulary. [3] [4] Signed acceptance became part of the process. [4] What the 2017 design lacked was a threat side to the ledger, which OMB itself identified. [4] Later direction, in EO 14028, added specific architecture goals, such as Zero Trust and secure cloud services, rather than more reporting. [6]

For any public body, the open question is where accepted risk is recorded. When it is kept in a security office alone, it is a warning. When it is kept beside the budget, signed by someone who controls the money, it is a decision. [1] [3]

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

  • Executive Order 13800
  • OMB M-17-25
  • risk acceptance
  • federal IT budget
  • NIST Cybersecurity Framework
  • shared services