An auditor picks one row from the risk spreadsheet: “customer data leak”, owner “IT”, rating “high”, last changed many months ago. They ask who exactly is accountable for this risk, what decision was taken and when anyone last checked that the assessment still holds. The spreadsheet answers none of these questions, even though the row is there and nobody disputes that the risk exists.
A risk register is the document that turns assessment results into decisions and lets you demonstrate them later. NIST IR 8286r1, published in December 2025, describes the cybersecurity risk register as a formal channel through which risk information reaches the people who decide on the risks of the whole enterprise. Whether that channel works depends on three things: the columns, the owner and the review rhythm.
What is a risk register?
A risk register is an organised record of identified risks together with their assessment, the decision on how to respond and the person accountable for them. NIST IR 8286r1 quotes the definition from the US Office of Management and Budget (OMB Circular A-11): a repository of risk information, including the data understood about risks over time.
The same publication cites the ISO definition, under which a risk register (or risk log) is a record of information about identified risks. Both definitions share one point: a register is not a one-off report, but a collection that changes along with the risks.
NIST recommends that each high-value system and each organisation should have its own cybersecurity risk register. At higher levels these registers are aggregated, normalised and prioritised into an enterprise risk register used by senior management. A register is therefore a tool of risk management, not risk management itself.
How does a risk register differ from risk analysis and risk assessment?
Risk analysis and risk assessment are processes, and a risk register is the document in which their results and the decisions taken are recorded. How IT risk analysis works and why cybersecurity risk assessment is the foundation of a security programme are covered in separate articles; the register takes over their output.
NIST IR 8286r1 separates two kinds of record. A risk assessment report, whose elements are described in Appendix K of NIST SP 800-30r1, documents the planning, execution and evaluation of risks. The register complements it: a short summary that can be read at a board meeting.
Details that do not fit in a row go into a risk detail record. According to NIST, the detail record holds the assumptions, considerations, people involved and schedules, so that the formal register remains a summary rather than a lengthy report. If the description of a single risk takes a page, that page belongs in the detail record, and the register keeps a reference to it.
Which columns should a risk register have?
The notional template in NIST IR 8286r1 has these columns: identifier, priority, risk description, risk category, current assessment (likelihood, impact, exposure rating), risk response type, risk response cost, risk response description, risk owner and status. NIST also quotes OMB, according to which a risk register typically contains a description of the risk, its impact, the probability of its occurrence, mitigation strategies, risk owners and a ranking to identify higher-priority risks.
| Column | What it contains | What to avoid |
|---|---|---|
| Identifier and priority | a permanent entry number and relative importance (an ordinal number or a level on a scale) | renumbering entries every time the list is sorted |
| Description and category | the risk scenario and a common theme that allows registers to be combined | a one-word description and categories invented for each entry |
| Current assessment | likelihood, impact and the resulting exposure | mixing pre-response and post-response risk in one column |
| Response | the response type, its description and estimated cost | a response with no cost and no deadline |
| Owner and status | the person accountable for the risk and the current state of the entry | a department instead of a person, and a status nobody updates |
NIST notes that, depending on the risk strategy, a register may have more or fewer fields, and some information may be held elsewhere as long as it is linked to the entry. Consistency matters more than the number of columns: the same model has to be used in every entry and in every cycle.
In the NIST model, the response type takes one of four values for negative risks: accept, transfer, mitigate or avoid. Acceptance applies to risks within tolerance and, according to NIST, requires no action other than monitoring.
How do you describe a risk so that the entry is clear?
You describe a risk as a scenario: what or who could act, through which weakness, against which asset and with what consequence for the business. For cybersecurity risks, NIST IR 8286r1 lists the elements of a complete scenario: a threat source, a threat event, a vulnerability the threat source might exploit, the assets that may be affected and the resulting harmful impact.
The NIST example shows how this reads in practice: construction work severs a critical fibre optic cable that was not protected in conduit, interrupting communications to the data centre and causing the loss of availability of financial systems. One sentence contains the cause, the weakness, the asset and the consequence, so the reader does not have to guess what the risk is about.
An entry reading “ransomware” carries none of that information. An entry reading “a criminal group encrypts the ERP servers through an unpatched VPN gateway, halting invoicing” immediately shows who should own the risk, which response makes sense and what needs to be measured in the assessment.
NIST also recommends common phrasing for descriptions, in which threats, assets, methods and impacts appear in every scenario. Without it, registers from different departments cannot be compared or combined.
How do you record the risk assessment in the register?
The register records the likelihood, the impact and the resulting exposure, and clearly marks whether the assessment refers to the risk before the response, after it, or to the target state. NIST IR 8286r1 calls the combination of impact and likelihood exposure and notes that ISO 31000 and NIST SP 800-30r1 use the term level of risk for it.
NIST distinguishes several assessments. Inherent risk describes conditions without any organisational intervention, current risk takes the steps already taken into account, and after the response the register may show residual risk. Some organisations also record target residual risk, the state the company aims for in line with its risk appetite and tolerance.
In the notional register in Appendix D of NIST IR 8286r1, a PC stolen from the reception area has a likelihood of 0.75 and an impact of 0.1, giving an exposure of 7.5% (low), an “accept” response and zero cost. The scales used to set such values, and the colour thresholds, are covered separately in the article on the cybersecurity risk matrix.
Record the date next to every assessment. An assessment without a date cannot tell a risk checked last week from one nobody has touched since the register was set up.
Who is the risk owner?
The risk owner is the designated party responsible and accountable for ensuring that the risk is maintained in accordance with the organisation’s requirements, which is how NIST IR 8286r1 defines the role. The owner may work with a risk manager who manages and monitors the selected response.
According to NIST, each risk in the register is assigned a risk owner, who uses the priority to select and assign appropriate responses while taking business objectives into account. The management responsible for enterprise risk should make sure that this accountability is real.
NIST draws attention to the difference between responsibility and accountability. In a US federal agency, responsibility for a system’s risks might lie with the system owner, while accountability lies with the authorising official. The risk strategy should clearly describe which roles are responsible and which are accountable for risk decisions at each level of the organisation.
A good rule: the owner should be someone who has influence over the budget and the process the risk concerns. The IT department can be a good risk manager, but the owner of the risk of invoicing coming to a halt should be someone responsible for invoicing. In the NIST notional register, the ransomware risk also lists a backup owner, which is worth copying into your own template.
Key takeaways:
- The risk owner is a specific person, not a department.
- The owner selects the response, and the risk manager runs and monitors it.
- A backup owner in the register prevents gaps in decision-making.
How often should you review the risk register?
The company sets the review frequency in its risk management strategy, and NIST IR 8286r1 recommends that it be written down together with the method and the person responsible. According to NIST, risk management programmes should specify the frequency and methods for monitoring, evaluating the effectiveness of and adjusting risk responses, and should name the bodies that discuss and approve the most significant risks.
The calendar is only part of the rhythm. NIST describes continuous monitoring that can show a risk approaching the risk tolerance level, triggering a review of the risk that may raise its priority. The limits such a trigger refers to are set by the risk appetite and tolerance agreed by the board.
According to NIST, the register should be monitored continually, including risks accepted as residual risk. Changes in the risk landscape, including changes in regulation, may require a periodic review of risk appetite, tolerance and key risk indicators.
Date fields help. Among the contents of a risk detail record, NIST lists the date on which the risk was first documented, the date of the last assessment, completion dates for mitigations and the date of the next expected assessment. A “next review” column in the register turns a general principle into a task list for specific weeks.
Does the law require you to keep a risk register?
Neither NIS2 nor the GDPR uses the term “risk register”, but both require risk management that you must be able to demonstrate. A register is the simplest way to give that demonstration a consistent form.
Legal position as of 23 October 2026. Article 21 of the NIS2 Directive requires Member States to ensure that essential and important entities apply risk-management measures that include, at a minimum, policies on risk analysis and information system security. The directive works through national law: an entity’s obligations arise from the provisions by which its Member State has transposed it.
Article 32 of the GDPR requires measures ensuring a level of security appropriate to the risk, and Article 24 calls for measures that are reviewed and updated where necessary. The accountability principle in Article 5(2) adds that the controller must be able to demonstrate compliance.
Standards are voluntary, but they bring order to the work. ISO/IEC 27005:2022 provides guidance on managing information security risks, and ISO 31000:2018 provides general risk management guidelines.
When should you close or archive a risk?
A risk is archived when it no longer concerns the company, and this is done on the basis of evidence, not by deleting the row. NIST IR 8286r1 recommends considering an annual review of risks for archiving and, at the same time, reviewing the archive for risks that re-emerge.
According to NIST, the review should demonstrate through documentation, with quantitative support where possible, that the risk has been overcome by events, modified by a change in the organisation’s direction, or effectively resolved through mitigation demonstrated over time. The point is to make sure the risk does not emerge again soon after archiving.
Closing is handled by the status column, which in the NIST template tracks the current condition of the risk and any subsequent activities. In the NIST notional register the statuses are “open”, “closed” and “updated”, and a risk transferred through insurance has the status “closed” but still appears in the register.
How do you build a risk register step by step?
You build a register from its rules, not from its entries: first common categories, scales, roles and review rhythm, then scenarios, assessments and decisions. The table sets out the steps with the evidence an auditor may ask for.
| Step | What is produced | Who | Evidence for the auditor | Typical mistake |
|---|---|---|---|---|
| 1. Setting the register rules | categories, scales, roles, review frequency | the board and the person responsible for risk | an approved, dated procedure | categories created for each entry |
| 2. Writing the scenarios | risk descriptions for critical systems and processes | process owners with the security team | risk assessment results | a one-word description |
| 3. Assessment on a common scale | likelihood, impact, exposure, priority | the risk analyst | an assessment report or risk detail record | current and residual risk in one column |
| 4. Appointing owners | the name of the owner and a backup | the board | confirmation that the role was accepted | a department instead of a person |
| 5. Response decision | response type, description, cost and deadline | the risk owner | a dated, signed decision | a response with no deadline or cost |
| 6. Review rhythm | dates of the last and next assessment, triggers | the person responsible for the register | the change history of entries | an annual review that ignores events |
| 7. Archiving review | an archive with reasons for closure | the risk committee or the board | review minutes | deleting rows without a trace |
Without steps 1 and 6, a register rarely survives its first year. Without rules, every department keeps its own version, and without a review rhythm the register ages faster than the risks it describes.
How can nFlo help you build a risk register?
In the governance, risk and compliance area, nFlo builds risk registers in two services. In the ISO/IEC 27005 - Information Security Risk Management service we prepare an information asset register with owners, a risk matrix, a risk treatment plan, a Statement of Applicability and a residual risk register with formal acceptance, together with a monitoring process with review triggers (new system, incident, organisational change) and a management reporting cycle.
In the ISO 31000 - Enterprise Risk Management service we build a centralised corporate risk register in which each risk has an identifier, an owner, an inherent and residual risk assessment and a treatment plan with deadlines and effectiveness metrics, together with training for risk owners. If your register cannot answer an auditor’s questions today, contact us.
What else is worth knowing about risk registers?
Can a risk register be kept in a spreadsheet?
Yes. NIST lists dashboards and GRC tools among the formats for risk register data, alongside spreadsheet files in CSV format. The format matters less than consistency: the same categories, scales and fields in every entry, and a change history that shows who changed an assessment and when.
How does a risk register differ from an asset register?
An asset register says what the company protects and who owns each asset. A risk register describes what could happen to those assets, how likely and severe it is, and what decision was taken. The first is the starting point for the second.
Who approves the risk register?
The most significant risks and response plans should be approved by the body named in the risk management strategy, for example a risk committee or the board. NIS2 requires Member States to ensure that the management bodies of essential and important entities approve their cybersecurity risk-management measures.
Does an accepted risk still need to be monitored?
Yes. According to NIST, acceptance means no further action except monitoring, and the register should be monitored continually, including risks accepted as residual risk. An accepted risk stays in the register with its owner and a next review date.
Does every risk in the register need an owner?
Yes. In the NIST model, each risk in the register is assigned a risk owner who is accountable for selecting responses in line with its priority. An entry without an owner is a risk nobody takes decisions about.
