M-17-05 started the seven-day notice to Congress at the agency’s determination, so deciding became the control
The budget office’s M-17-05 of November 4, 2016 defined a federal major incident by likely harm, with one fixed line at 100,000 people’s personal information. Its seven-day notice to Congress ran from the agency’s determination, making the decision process the control to watch.
Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.
Two-minute summaryThe dek and the first sentence of each section, quoted from the article
The budget office’s M-17-05 of November 4, 2016 defined a federal major incident by likely harm, with one fixed line at 100,000 people’s personal information. Its seven-day notice to Congress ran from the agency’s determination, making the decision process the control to watch.
- The week The memorandum’s first section addressed reporting.
- The standard: likely harm, judged by the agency The same words appear in PPD-41.
- The bright line: 100,000 people The memorandum set one rule that did not depend on a harm forecast.
- The clock: seven days from a determination M-17-05 restated the duty with a change of emphasis.
- The deciders: who decides and what follows The memorandum named who must take part.
- How the memorandum reached different agencies A cabinet department with its own security operations center could gather facts quickly.
- Objections, in order OMB’s guidance was meant to implement the statute, and an agency acting in good faith would decide promptly. [1] [4] But a control should work when people are cautious or slow, as well as when they are prompt.
- What an agency CISO and general counsel could check today The procedure can set internal deadlines from detection to determination.
- With hindsight OMB revised its fiscal-year FISMA guidance in later years, and the major incident definition carried forward.
The week
On November 4, 2016, the Office of Management and Budget issued M-17-05, Fiscal Year 2016 - 2017 Guidance on Federal Information Security and Privacy Management Requirements. [2] The archived list of the office’s memoranda gives that date and shows the memorandum at 12 pages. [2]
The memorandum’s first section addressed reporting. It said quarterly and annual security metrics serve to ensure agencies implement Administration priorities and cybersecurity practices, and to supply data for oversight. [1] It said the FY 2016 metrics were aligned to the five functions of NIST’s Cybersecurity Framework: Identify, Protect, Detect, Respond and Recover. [1]
It set deadlines. All federal agencies, including small and independent agencies, were to report against the annual FY 2016 metrics by November 10, 2016. [1] Agency annual reports were due to Congress by March 1, 2017. [1] Agencies were to report cybersecurity performance through DHS’s CyberScope system. [1]
The memorandum also said agencies would continue to move toward automated data collection and a federal Continuous Diagnostics and Mitigation dashboard, which would begin replacing the existing data collection process. [1]
Its second section was titled Updated Major Incident Definition and DHS US-CERT Incident Notification Guidelines. [1] It said the Federal Information Security Modernization Act of 2014 authorizes the budget office to define the term major incident and directs agencies to notify Congress of one. [1] It said the guidance replaced the definition in M-16-03, the guidance for fiscal years 2015 and 2016. [1]
The new definition was based on harm. A major incident was any incident likely to result in demonstrable harm to the national security interests, foreign relations or economy of the United States. [1] Harm to the public confidence, civil liberties, or public health and safety of the American people also counted. [1]
To judge impact, agencies were to use the incident management process in NIST Special Publication 800-61. [1] They were encouraged to use the National Cybersecurity Incident Scoring System, which weighs eight factors, including functional impact, information impact, recoverability, cross-sector dependency and potential impact. [1] A footnote tied the definition to Level 3 or higher on the Cyber Incident Severity Schema. [1]
The memorandum set separate terms for breaches of personal information. A breach was a major incident when compromise of the information involved was likely to cause the same demonstrable harm. [1] It also set a fixed line. Unauthorized modification, deletion, exfiltration of or access to the personal information of 100,000 or more individuals constituted a major incident. [1]
It set the timing. Agencies must notify the appropriate congressional committees of a major incident no later than seven days after the date on which the agency determined that it had a reasonable basis to conclude one had occurred. [1] The memorandum’s requirements tracker added that DHS would notify the budget office within one hour of an agency notifying DHS of a major incident. [1]
It linked the definition to a directive issued that July. Other than breaches, an incident meeting the major incident definition was also a significant cyber incident under Presidential Policy Directive 41. [1] PPD-41, dated July 26, 2016, set out how the federal government would coordinate its response to significant cyber incidents. [3]
M-17-05’s title describes an annual reporting memorandum. [2] Its second section did something with longer reach. It defined when a federal agency must tell Congress about a cyber incident, and it set when the telling must begin. [1]
The definition rested on likely harm, judged by the agency. [1] The clock ran seven days from the date on which the agency determined it had a reasonable basis to conclude a major incident had occurred. [1] Only one part of the definition was automatic: a breach touching the PII of 100,000 or more people. [1]
Our argument is that M-17-05 made the agency’s own determination the hinge of federal incident reporting, so the process for deciding became the control to examine. We take the memorandum in four parts:
- The standard: how an agency judged whether an incident was likely to cause demonstrable harm.
- The bright line: the one count that made a breach major without further judgment.
- The clock: when the seven days began and what else ran alongside them.
- The deciders: who the memorandum said must take part, and what the label set in motion.
The parts do not overlap. The first is the judgment, the second the exception to judgment, the third the timing and the fourth the people and consequences. After them, we look at how the memorandum reached different agencies and their contractors, take the objections in order, and end with what an agency CISO and general counsel could check today.
This is a reading of public documents. Agencies should confirm current requirements with OMB guidance in force and their counsel.
The standard: likely harm, judged by the agency
The words OMB chose
The definition’s core phrase was likely to result in demonstrable harm. [1] The harms listed were to national security interests, foreign relations or the economy of the United States, or to public confidence, civil liberties, or public health and safety. [1]
The same words appear in PPD-41. The directive defines a significant cyber incident as one likely to result in demonstrable harm to the same list of interests. [3] In our reading, OMB aligned the federal agency definition with the national one so that one judgment could serve both purposes.
A likelihood, not an outcome
The test asked whether harm was likely, not whether it had happened. [1] That makes the determination a forecast. An agency would need to judge, early in an incident, what the attacker could do with what it reached.
A forecast is harder to audit than an outcome. Two agencies with similar incidents could forecast differently and both be acting in good faith. The memorandum offered tools to narrow that range.
The tools for judging
The memorandum pointed to NIST SP 800-61, the Computer Security Incident Handling Guide, for the incident management process. [1] It encouraged use of US-CERT’s National Cybersecurity Incident Scoring System. [1]
The scoring system’s eight factors were functional impact, observed activity, location of observed activity, actor characterization, information impact, recoverability, cross-sector dependency and potential impact. [1] The memorandum said using it created uniformity in terminology and criteria between agencies and US-CERT responders. [1]
The severity schema
A footnote anchored the definition to the Cyber Incident Severity Schema. It said a major incident was Level 3, orange, or higher. [1] It quoted Level 4, red, as likely to result in a significant impact to public health or safety, national security, economic security, foreign relations or civil liberties. [1] It quoted Level 5, black, as posing an imminent threat to wide-scale critical infrastructure services, national government stability or the lives of US persons. [1]
The footnote did not quote Level 3’s definition. [1] In our reading, that left the threshold level described only by the main text’s phrase, demonstrable harm. A scoring tool helps an agency place an incident on a scale. Where the line falls on that scale stays a judgment.
Each factor and the record behind it
The scoring factors read like a list of records an agency should already keep. [1] Functional impact asks what the affected system does for the mission, which the system owner’s documentation should say. Observed activity and its location ask what the attacker did and where, which logs should show.
Actor characterization asks who the attacker appears to be, which may depend on threat intelligence the agency receives from others. Information impact asks what data was involved, which a data inventory should answer. Recoverability asks whether the agency can restore operations and data, which tested backups and contingency plans should show. [1]
Cross-sector dependency asks whether others depend on the system. Potential impact asks what could happen next. [1] Those two are forecasts, and they draw on the other six. An agency strong on records can score the first six quickly and spend its judgment on the last two.
In our reading, this is why the definition, though written as a harm test, worked in operation as a records test. The quality of an agency’s inventories and logs set how fast it could reach a reasoned determination.
Consistency across agencies
The memorandum said using the scoring system created uniformity in terminology and criteria between agencies and US-CERT responders. [1] Uniform inputs do not guarantee uniform outcomes, but they make differences visible. Two agencies scoring similar incidents differently can be asked why, factor by factor.
That makes the scores themselves records worth keeping. An agency that documents its factor scores alongside its determination can show how it reached the answer, and an overseer can compare it with others.
What the standard asked of the agency
To apply the standard, an agency needs facts quickly: what systems were reached, what data, what the actor appears to be and whether the agency can recover. Those are the scoring factors. [1] An agency that cannot gather them early cannot make a reasoned forecast.
In our reading, the standard therefore rewarded agencies with good asset inventories, logging and data maps. An agency that did not know what a compromised server held could not judge information impact. Its determination would wait on discovery, and the clock would wait with it.
Below the threshold
Incidents that did not meet the definition still had to be handled. The memorandum’s requirements tracker shows US-CERT providing agencies a quarterly log of incidents they reported, which agencies would validate. [1] The tracker also lists reporting in a revised US-CERT incident reporting format, with an April 1, 2017 deadline. [1]
Those items show that the major incident label sat on top of a routine flow of incident reports to US-CERT. The label decided what went to Congress. It did not decide whether an incident was reported to DHS at all.
The bright line: 100,000 people
What it said
The memorandum set one rule that did not depend on a harm forecast. An unauthorized modification of, deletion of, exfiltration of or access to 100,000 or more individuals’ PII constituted a major incident. [1]
It defined each act. Modification was changing components of information or systems. [1] Deletion was removing information from a system. [1] Exfiltration was obtaining information without authorization or beyond authorized access, without modifying or deleting it. [1] Access was logical or physical access without permission to federal information or systems. [1]
Access was enough
The fourth act is the widest. Unauthorized access to the records of 100,000 people met the line, whether or not data left the network. [1] An agency could not wait to prove exfiltration. If an attacker reached a database of that size, the count was met.
That turned a technical question into a counting question. How many distinct people’s PII sat in the systems the attacker could reach. An agency with a data inventory could answer it. An agency without one would struggle, and its uncertainty would delay the determination.
Counting distinct people
The line counted individuals, not records. [1] A database can hold several records for one person, and one person can appear in several databases. An agency would need to deduplicate across the systems the attacker reached to know whether 100,000 people were affected.
That is a data engineering task. It needs a common identifier or a matching method, and it needs to run fast. An agency that had never counted distinct people across its systems would be doing it for the first time during an incident.
Integrity counts too
Two of the four acts concern integrity. Unauthorized modification and unauthorized deletion of 100,000 people’s PII each met the line. [1] An attacker who altered benefit records, without copying them, would trigger it.
That widened the line beyond data theft. Detecting modification needs integrity monitoring or reconciliation against a trusted source. An agency that watched for exfiltration alone could miss the integrity half of its own bright line.
Why a bright line
A bright line removes argument. Two agencies with the same count reach the same answer. Congress gets notice of large breaches without depending on each agency’s forecast of harm. The memorandum still kept the harm test for breaches below the line. [1]
The memorandum gave no reason for choosing 100,000 in the text we could read. [1] Its effect is clearer than its rationale: above the line, the agency counts; below it, the agency forecasts.
What the line did not count
The line counted people, not sensitivity. [1] A breach of 99,000 records containing security clearance files would fall to the harm test. A breach of 100,000 records holding names and work email addresses would meet the line automatically. The harm test could still catch the first. The line could over-include the second.
The memorandum separated two assessments. It said analysis for reporting a major breach to Congress is distinct from assessing the risk of harm to individuals, for which agencies should use OMB’s breach guidance. [1] Congressional reporting and notice to affected people ran on separate tracks.
Breaches and PPD-41
The memorandum treated breaches differently for coordination purposes. A breach of PII that was a major incident met the definition of a significant cyber incident under PPD-41 only when it resulted from a cyber incident. [1] A lost box of paper records could be a major incident for Congress without triggering the PPD-41 coordination structure. [1]
The clock: seven days from a determination
What the statute said
FISMA 2014 directs an agency to notify the appropriate congressional committees of a major incident. [4] The notice is due not later than 7 days after the date on which there is a reasonable basis to conclude that the major incident has occurred. [4] The same law directed OMB to develop guidance on what constitutes a major incident. [4]
What OMB wrote
M-17-05 restated the duty with a change of emphasis. Agencies must notify Congress no later than seven days after the date on which the agency determined that it had a reasonable basis to conclude that a major incident occurred. [1] A footnote said that once an agency, based on initial analysis, arrives at a reasonable basis to conclude a major incident occurred, it must report within seven days. [1]
In our reading, the statute’s wording is objective: a reasonable basis exists or it does not. OMB’s wording ties the start to the agency’s determination. The difference matters if an agency is slow to decide. Under OMB’s wording, the clock waits for the decision. Under the statute’s, an argument exists that it began when the facts were there.
What a reasonable basis requires
The memorandum’s footnote tied the reasonable basis to initial incident analysis. [1] In our reading, a reasonable basis is less than proof. It is enough evidence that a careful official would conclude a major incident probably occurred.
That reading has an operational consequence. An agency need not wait for a forensic report to decide. It can decide on the initial analysis and correct the picture in supplements, which the memorandum required anyway. [1] Agencies should confirm with counsel how they apply the standard.
Why that makes the decision process the control
If the clock starts at the determination, the speed and quality of the determination become what an overseer should test. A fast, documented decision process produces timely notice. A slow or undocumented one produces late notice that still looks compliant on its face.
An inspector general or a congressional committee reviewing an incident would ask two questions. When did the agency have the facts that supported a reasonable basis? When did it decide? The gap between the two is where the memorandum’s design puts the risk.
A hypothetical sequence
A hypothetical sequence shows where the risk lies. A security analyst sees unusual queries against a benefits database. The incident response team confirms an intruder and finds the account used. The data team works out which tables the account could read and counts the people in them.
The privacy official reviews what fields were exposed. The CISO scores the incident. The CIO and system owner join a call, and the designated official decides the incident is major. Under OMB’s wording, the seven days begin at that decision. [1]
In this hypothetical, the data team’s count may decide the matter on its own, if it reaches 100,000 people. [1] Every step before the decision takes time, and none of it is visible to Congress. A review after the fact would want each step dated. That record is what shows whether the agency decided as soon as it had a reasonable basis.
Routine reporting before the label
The major incident decision does not replace routine incident reporting to DHS. The requirements tracker refers to agencies reporting incidents to US-CERT and validating the quarterly log US-CERT returns. [1] An incident would be reported to US-CERT as an incident well before anyone called it major.
That gives an overseer a second date to compare. The date an incident was first reported to US-CERT, set against the date it was determined major, shows how long the agency took to reach its conclusion.
What the first report should hold
The memorandum said the initial report should take into account the information known at the time, the sensitivity of the details and the classification level of the information. [1] It did not require certainty. A reasonable basis to conclude was enough to start the clock, and the report could say what was not yet known. [1]
The supplements
The memorandum required more after the first report. Once an incident was identified as major, the agency must give Congress, as soon as available, additional information on the threats, actors and risks posed. [1] It listed previous risk assessments of the affected system, its current status, and the detection, response and remediation actions taken. [1]
Those supplements need records. A risk assessment done before the incident must exist to be provided. Remediation actions must be logged to be described. In our reading, the supplement requirement reached back into an agency’s routine security documentation.
The one-hour link
The requirements tracker set a shorter clock for a different actor. DHS was to notify OMB within one hour of an agency notifying DHS that a major incident had occurred. [1] That item placed DHS between the agency and OMB, so OMB learned of a major incident through DHS. [1]
When facts change after the decision
An initial determination rests on initial analysis. [1] Later facts may show the incident was larger, or smaller, than first thought. The memorandum required supplements as information became available. [1] It did not describe withdrawing a determination in the text we could read. In our reading, an agency that finds an incident was smaller should say so in a supplement rather than leave the first report uncorrected.
Notice to individuals
The tracker also restated the duty to notify affected individuals, in accordance with FISMA 2014, as expeditiously as practicable and without unreasonable delay. [1] That duty had no fixed number of days. It ran on its own track, separate from the seven-day congressional clock. [1]
The deciders: who decides and what follows
The people in the room
The memorandum named who must take part. Appropriate analysis of an incident would include the agency CIO, the chief information security officer, mission or system owners, and, if the incident was a breach, the senior agency official for privacy. [1]
Each brings different facts. The CISO knows the technical scope. The system owner knows what the system does and what its loss would mean. The privacy official knows what PII is involved and what harm it could cause. The CIO can commit resources. The harm forecast needs all four.
Who holds the pen
The memorandum named participants but not a single decision maker. [1] In our reading, an agency needed to decide for itself who signs the determination, and how disagreement is resolved. Without that, the four participants could each wait on the others.
The privacy official’s role
The senior agency official for privacy joins the analysis when the incident is a breach. [1] For the bright line, the privacy official’s question is how many people. For the harm test, it is what the exposed information could do to them and to public confidence.
The memorandum separated the congressional analysis from the assessment of harm to individuals, which follows OMB’s breach guidance. [1] The privacy official may run both at once, for different audiences and on different clocks.
Recording a decision not to call it major
The memorandum’s clock runs from a determination that an incident is major. [1] A determination that an incident is not major has no clock and no audience. Yet it is a decision that will be questioned if the incident grows.
A record of that decision, with the facts known, the scores and the reasons, protects the agency and informs the next review. The memorandum did not require such a record in the text we could read. [1] In our reading, an agency that wants its process to be testable would keep one anyway.
What the label triggers
The label had consequences beyond Congress. Other than breaches, a major incident was also a significant cyber incident under PPD-41, and would trigger the directive’s coordination mechanisms, including a Cyber Unified Coordination Group. [1]
PPD-41 described what that meant. It said a Cyber Unified Coordination Group serves as the primary method for coordinating among federal agencies in response to a significant cyber incident. [3] It set three concurrent lines of effort: threat response, asset response, and intelligence support. [3]
The directive named the leads. The Justice Department, acting through the FBI and the National Cyber Investigative Joint Task Force, would lead threat response. [3] DHS, through its National Cybersecurity and Communications Integration Center, would lead asset response. [3] The Office of the Director of National Intelligence would lead intelligence support. [3]
The agency inside the coordination group
Under PPD-41, the affected agency is not the lead for threat response or asset response. [3] Threat response, led by the Justice Department through the FBI and the National Cyber Investigative Joint Task Force, includes investigation, evidence collection and attribution. [3] Asset response, led by DHS, includes technical assistance to protect assets, mitigate vulnerabilities and reduce impacts. [3]
PPD-41 gives the affected federal agency its own line of effort. It says the agency manages the incident’s impact, including continuity, privacy, legal and regulatory compliance such as notification, and congressional inquiries. [3] It says the affected agency has primary responsibility for that line. [3] The congressional notice under FISMA falls within it.
That division adds coordination work at the moment the agency has least capacity. In our reading, an agency that has practiced the hand-offs to the FBI and DHS before an incident will spend less of its scarce time on them during one.
Why consequences shape the decision
A label that brings congressional notice and an interagency coordination group is a heavy label. An agency might hesitate to apply it. The memorandum’s design depended on agencies applying it promptly anyway. [1]
In our reading, that is the deepest reason the decision process is the control. An agency that fears the consequences of the label has a reason to delay the determination, and OMB’s wording let the clock wait for it. A documented process with deadlines for each step removes that room.
Contractors and systems operated for agencies
FISMA makes agency heads responsible for protecting information systems used or operated by an agency or by a contractor of an agency on its behalf. [4] An incident at a contractor that held an agency’s data could therefore be the agency’s major incident. The agency’s determination would then depend on facts the contractor held.
That puts the decision process partly in the contract. An agency needs its contractors to report quickly and in enough detail to let it judge harm and count records. The memorandum did not set contractor terms. [1] That was left to agencies and their contracting officers.
How the memorandum reached different agencies
Large departments
A cabinet department with its own security operations center could gather facts quickly. Its difficulty was coordination among components, each with its own CIO and system owners. The memorandum’s list of participants applied to the agency, which for a department means deciding who participates at headquarters. [1]
Components that run their own systems
A department whose components run their own networks and security teams faces a question the memorandum left to it. Does the component decide, or headquarters? The memorandum named the agency CIO and CISO among the participants. [1] A department would need to say whether that means the departmental officials, the component’s, or both.
If a component decides and headquarters learns late, the department’s seven days may be partly spent before its own officials engage. Writing down the escalation path from component to department is part of the decision process the memorandum made central.
Small and independent agencies
The memorandum’s reporting duties reached all federal agencies, including small and independent agencies. [1] A small agency may rely on a shared service provider or DHS services for monitoring. Its facts would come from outside, and its determination would depend on how fast they arrived.
Agencies with large PII holdings
An agency whose single benefits, tax or personnel database holds records on more than 100,000 people would meet the bright line if an attacker reached it. [1] For them, the count was not the hard part. The hard part was knowing quickly whether an attacker had reached that database at all.
Inspectors general
Inspectors general reported on agency programs under FISMA through the IG metrics the memorandum referred to. [1] An IG could use the major incident process as a test of the agency’s program: whether the determination was timely, documented and consistent with the definition.
Contractors and cloud providers
A contractor operating a system for an agency would not report to Congress itself. Its role was to give the agency the facts. [4] In our reading, a contractor that understood the bright line and the harm factors could report in a form the agency could use within its seven days. A contractor that reported only that an incident had occurred, without the systems reached or the number of people affected, would leave the agency unable to decide. The agency’s determination can be no faster than its slowest source of facts.
Shared service providers
An agency that provides shared services to others, such as payroll or hosting, presents a further case. An incident at the provider could touch data belonging to several customer agencies. Each would face its own determination for its own data. The memorandum addressed agencies one at a time. [1] A provider and its customers would need to agree in advance how facts are shared so each can decide within its own clock.
The committees on the receiving end
The memorandum listed the committees to be notified. They were the House Committees on Oversight and Government Reform, Homeland Security, and Science, Space, and Technology. [1] On the Senate side, they were the Committees on Homeland Security and Governmental Affairs and on Commerce, Science, and Transportation, plus the appropriate authorization and appropriations committees. [1]
Several committees receive the same notice. An agency’s legislative affairs office therefore becomes part of the reporting chain, and its procedures need to fit inside the seven days.
Objections, in order
1. The determination wording is harmless
OMB’s guidance was meant to implement the statute, and an agency acting in good faith would decide promptly. [1] [4] But a control should work when people are cautious or slow, as well as when they are prompt. Wording that lets the clock wait gives caution room. A process with step deadlines closes it.
2. A harm test is too vague
It is a judgment. The memorandum narrowed it with a scoring system and a severity schema. [1] The alternative, a list of incident types, would miss new kinds of harm. The bright line for PII gave certainty where a count was possible. [1]
3. The 100,000 line is arbitrary
Any line is. The memorandum kept the harm test for breaches below it, so a smaller sensitive breach could still be major. [1] The line traded precision for speed and consistency at the high end.
4. Seven days is too short to know anything
The memorandum anticipated that. It said the first report should reflect what was known at the time, and required supplements as information became available. [1] Seven days is enough to say what happened so far.
5. Congress is the wrong first audience
Congress was the audience the statute named. [4] Operational coordination ran through DHS and, for significant incidents, PPD-41. [1] [3] The congressional notice served oversight, not response. In our view, keeping those purposes apart is a strength of the design.
6. Agencies will over-report to stay safe
Some may. Over-reporting has costs: congressional attention, PPD-41 coordination and staff time. [1] [3] The harm test and scoring factors are meant to reduce both over- and under-reporting. [1] A documented determination process protects an agency in both directions, because it shows why an incident was or was not called major.
7. Notice to Congress risks exposing sensitive details
The memorandum anticipated that concern. It said the report should take into account the sensitivity of the details and the classification level of the information. [1] An agency can notify within seven days without disclosing everything it knows. The duty is to notify, with what is known and appropriate to share. [1] [4]
What an agency CISO and general counsel could check today
A written determination procedure
The first check is whether the agency has a written procedure for deciding whether an incident is major, with named participants matching OMB’s list and a named signer. [1] [5]
Deadlines for each step
The procedure can set internal deadlines from detection to determination. The CISO gathers scoring facts by one point, the privacy official assesses PII by another, the signer decides by a third. The deadlines are the agency’s own; OMB guidance did not set them. [1] [5]
A record count on demand
The agency can test whether it can count, quickly, the distinct people whose PII is held in a given system. The bright line depends on that count. [1] [5]
Contractor reporting terms
Contracts for systems holding agency data can require reports in a form that supports the harm factors and the PII count. [4] The agency’s seven days begin from its own determination, but the facts it needs may sit with the contractor. [1]
A review of past determinations
Finally, the CISO and counsel can review recent incidents. For each, they can record when the facts supporting a reasonable basis were known and when the determination was made. The gap is the measure of the control. [1] [4]
An exercise of the procedure
A tabletop exercise can walk a realistic incident through the determination procedure, timing each step. The exercise shows where facts arrive late and where participants wait on each other. [1] [5] It also tests whether the agency can count distinct people across systems under time pressure. [5]
Alignment with breach response
The congressional determination and the breach response for individuals are separate analyses under OMB guidance. [1] [5] A check confirms that both procedures name the same privacy official and use the same count of affected people, so the agency does not tell Congress one number and the public another.
With hindsight
OMB revised its fiscal-year FISMA guidance in later years, and the major incident definition carried forward. On January 15, 2025, OMB issued M-25-04, Fiscal Year 2025 Guidance on Federal Information Security and Privacy Management Requirements, which rescinded M-24-04. [5]
M-25-04 defines a major incident in two branches. One is an incident likely to result in demonstrable harm to the national security interests, foreign relations or economy of the United States, or to public confidence, civil liberties, or public health and safety. [5] The other is a breach of PII likely to cause the same harm if compromised. [5] The harm language is the same as in 2016. [1] [5]
M-25-04 keeps the bright line. It requires a determination of major incident for any unauthorized modification, deletion, exfiltration of or access to the PII of 100,000 or more people. [5] It adds that other factors may also lead an agency to find a breach major. [5]
The 2025 guidance changed who is told and how fast. Agencies must report to CISA and the OMB Federal CIO’s office within 1 hour of determining that a major incident occurred. [5] They must notify the appropriate congressional committees and their Office of Inspector General no later than seven days after determining they have a reasonable basis to conclude a major incident occurred. [5]
M-25-04 also addressed the timing question directly. A footnote says the one-hour reporting is limited to the time after a major incident determination is made, not the detection of the incident. [5] It says an agency is expected to take some time to determine whether an incident reaches the major threshold. [5] In our reading, OMB’s guidance continued to start its clocks at the agency’s determination. [1] [5]
The 2025 guidance kept the link to PPD-41. It says that if a cyber incident is a major incident, it is also a significant cyber incident, triggering PPD-41’s coordination mechanisms and possibly a Cyber Unified Coordination Group. [5] It also kept the list of participants: the CIO, the CISO, mission or system owners and, for a breach, the senior agency official for privacy. [5] It added that agencies may consult OMB and CISA to make a determination. [5]
For incidents generally, M-25-04 requires agencies to report to CISA under CISA’s federal incident notification guidelines. [5] It says that includes events under investigation for 72 hours without a determination of their root cause or nature. [5] In our reading, that requirement puts a time limit on uncertainty for routine reporting to CISA, though not on the major incident determination itself. [5]
As checked on November 16, 2026, OMB’s list of memoranda on the White House website shows memoranda through M-27-01, dated October 23, 2026. [6] We did not find a later fiscal-year FISMA guidance memorandum in that list. [6] M-25-04 remains the latest such guidance we found. [5] [6]
What held up from November 2016 was the definition. The harm test, the 100,000-person line and the seven-day notice to Congress all appear in the 2025 guidance. [1] [5] What changed was the reach of notice, which now includes CISA, OMB and inspectors general on stated timelines. [5] What did not change was the starting point: the agency’s own determination. [1] [5]
Spotted an error? Tell us and we will check it and correct the article where needed.