GRC Cybersecurity: The Complete Guide
What GRC in cybersecurity means, what NIS2, DORA and NIST CSF 2.0 ask of leadership, who owns what, how to measure the program, and a build checklist.
GRC cybersecurity is the practice of running security as a governed business program instead of a collection of tools. Governance decides who owns cyber risk and how much of it the organization will carry. Risk management ranks what threatens your objectives and assigns treatment. Compliance proves to regulators, auditors and customers that the chosen controls actually work.
In Europe this has moved from good practice to legal duty. NIS2 requires management bodies to approve and oversee cybersecurity risk-management measures, and DORA gives a financial entity's management body the ultimate responsibility for ICT risk. This guide is the hub for every GRC page in the ResilientX Learning Center: what each discipline does, what the regulations ask, who owns what, how to measure the program, and a checklist to build one.
What does GRC mean in cybersecurity?
The term comes from the Open Compliance and Ethics Group (OCEG), which states that it originated the acronym in 2002. OCEG defines GRC as "the integrated collection of capabilities that enable an organization to reliably achieve objectives" while addressing uncertainty and acting with integrity.
In a security program, each letter has a concrete job:
- Governance is the decision system: who sets cyber strategy, who approves policy, who may accept a risk, and how leadership learns what is happening.
- Risk management is the prioritization system: which threats to which assets matter most, what treatment each gets, and what remains afterward.
- Compliance is the proof system: which external obligations apply, which controls meet them, and what evidence shows those controls run.
A security team can be technically excellent and still fail all three. If nobody with authority has accepted the residual risk of an unpatched legacy system, that is a governance gap. If the patching backlog is worked in ticket order instead of by exposure, that is a risk management gap. If the team patches well but cannot show an auditor when and how, that is a compliance gap.
For a plain-language introduction to the three letters outside security, see what GRC is. This guide stays with cybersecurity, where the term is most often used today and where European law now attaches duties to it.
Why is cyber GRC now a board-level obligation?
Three documents moved cyber governance out of the security team and into the boardroom.
NIS2. Directive (EU) 2022/2555, Article 20(1), requires the management bodies of essential and important entities to approve the cybersecurity risk-management measures taken under Article 21 and to oversee their implementation, and it makes them liable for infringements. Article 20(2) requires members of those management bodies to follow training. Member States had to transpose the directive by 17 October 2024 under Article 41(1).
DORA. Regulation (EU) 2022/2554, Article 5(2)(a), states that the management body of a financial entity bears the ultimate responsibility for managing its ICT risk. The rest of Article 5(2) makes the same body set roles, approve the digital operational resilience strategy including risk tolerance, approve ICT audit plans, allocate budget, and approve the policy on ICT third-party providers.
NIST CSF 2.0. The NIST Cybersecurity Framework 2.0 (NIST CSWP 29, published 26 February 2024) added GOVERN as a sixth Function next to IDENTIFY, PROTECT, DETECT, RESPOND and RECOVER. Subcategory GV.RR-01 states that organizational leadership is responsible and accountable for cybersecurity risk.
The financial exposure is explicit too. Under NIS2 Article 34(4), Member States must set maximum fines for essential entities of at least EUR 10 million or 2% of total worldwide annual turnover, whichever is higher. For important entities, Article 34(5) sets at least EUR 7 million or 1.4%.
The practical consequence for a CISO is a change of output. The program now has to produce decisions leadership can take and evidence that those decisions were taken, not only technical controls.
How do governance, risk and compliance fit together?
Treat them as one loop with three outputs, not three departments.
Governance sets objectives, appetite and decision rights. Risk management uses those limits to rank and treat risks. Compliance takes the controls chosen as treatment, maps them to obligations and tests them. Test failures and audit findings flow back into the risk register as new or re-scored risks, and anything above appetite goes back to governance for a decision.
When one of the three runs on its own, the loop breaks and the same issue gets argued in three meetings with three different spreadsheets.
| Governance | Risk management | Compliance | |
|---|---|---|---|
| Core question | Who decides, within what limits? | What matters most, and what do we do about it? | Can we prove the controls work? |
| Typical owner | Board, executive committee, CISO | Risk owners, coordinated by the CISO or risk function | GRC or compliance lead, control owners |
| Main artifacts | Charter, policies, appetite statement, board reports | Risk register, treatment plans, exceptions | Control library, framework mappings, evidence, audit reports |
| Cadence | Quarterly reporting, annual review | Continuous, with periodic reassessment | Control test cycle and audit calendar |
| Typical failure | Policies nobody enforces | Scores nobody acts on | Evidence collected the week before the audit |
The middle column in depth (ownership, appetite and a worked register entry from inherent to residual risk) is covered in our guide to GRC risk management.
What do NIS2, DORA and NIST CSF 2.0 ask of a GRC program?
The three texts use different words for the same building blocks. Mapping them once saves you from running three parallel programs.
| Building block | NIS2 (Directive 2022/2555) | DORA (Regulation 2022/2554) | NIST CSF 2.0 |
|---|---|---|---|
| Leadership accountability | Art. 20(1): management bodies approve, oversee and can be held liable | Art. 5(2)(a): management body bears ultimate responsibility for ICT risk | GV.RR-01: leadership responsible and accountable |
| Leadership competence | Art. 20(2): management body members follow training | Art. 5(2)(g): budget covers training and staff skills | GV.RR: roles, responsibilities and authorities |
| Appetite and tolerance | Art. 21(1): measures appropriate and proportionate to risk | Art. 5(2)(d): strategy includes the ICT risk tolerance | GV.RM-02: appetite and tolerance statements |
| Documented risk framework | Art. 21(2)(a): policies on risk analysis and system security | Art. 6(1): sound, comprehensive, well-documented ICT risk management framework | GV.RM-03: cyber risk included in enterprise risk management |
| Policy | Art. 21(2)(a), (h), (i): risk analysis, cryptography, access control and asset policies | Art. 5(2)(b): policies for availability, authenticity, integrity and confidentiality | GV.PO: policy established, communicated and enforced |
| Effectiveness review | Art. 21(2)(f): assess the effectiveness of measures | Art. 6(5): review at least once a year; Art. 6(6): internal audit | GV.OV-01: outcomes reviewed to adjust strategy |
| Supply chain | Art. 21(2)(d): supply chain security | Art. 5(2)(h): policy on ICT third-party providers | GV.SC: supply chain risk management |
| Incidents | Art. 21(2)(b): incident handling; Art. 23(4): early warning in 24 hours, notification in 72 hours | Art. 5(2)(e): approve response and recovery plans | RESPOND Function |
Two patterns stand out. First, every text puts leadership approval and periodic review at the center, so the board paper and the annual review are the highest-value artifacts you will produce. Second, the requirements overlap heavily, which is why a control should be assessed once and mapped to all three instead of being tested three times.
If NIS2 is your first obligation, our NIS2 compliance page shows how the requirements map inside the platform.
Who does what in a cyber GRC program?
Ownership is where most programs fail quietly. "Assign roles" is not enough; this is a split that works in practice.
| Role | Accountable for | Typical decisions |
|---|---|---|
| Board or management body | Approving cyber risk measures and appetite; overseeing implementation | Accept or reject residual risk above appetite; approve strategy and budget |
| CISO | Designing and running the security program; reporting to the board | Propose treatments; escalate risks above appetite; set control standards |
| Risk owners (business leaders) | The risks in their area and the treatment decision | Fund, accept or transfer a specific risk |
| Control owners | Operating controls and producing evidence | Fix failed controls; request exceptions |
| GRC or compliance lead | Framework mappings, test calendar, evidence and audit readiness | Schedule tests; chase exceptions to closure |
| Internal audit | Independent assurance that the framework works | Report findings to the board |
Two rules keep the split honest. The person who runs a control should not be the only person who tests it. And nobody below the agreed authority level can accept a risk above appetite, however urgent the release.
In a small company several of these roles sit with one person. That is fine, as long as the decisions are still written down and the approval for risk above appetite still comes from leadership.
What are the core processes of cyber GRC?
Seven processes cover almost everything regulators and auditors ask about.
- Policy lifecycle. Write, approve, publish, collect acknowledgments, review. Each policy has an owner and a review date.
- Risk assessment and treatment. Identify, score inherent risk, choose a treatment, score residual risk, compare it with appetite.
- Control design and testing. Each control has an objective, an owner, a frequency and a defined piece of evidence. Tests follow a calendar, not the audit date.
- Exceptions and risk acceptance. Any deviation from policy gets written acceptance from someone with the authority to give it, plus an expiry date.
- Third-party oversight. Suppliers are tiered by criticality and assessed accordingly, often with a standard questionnaire such as the SIG questionnaire.
- Incident governance. Classification, escalation and regulatory reporting. Under NIS2 Article 23(4), a significant incident needs an early warning within 24 hours and an incident notification within 72 hours of becoming aware of it.
- Assurance and reporting. Internal audit, management review, and a board report that shows risk against appetite instead of activity counts.
Processes 2, 3 and 4 are where the loop either closes or breaks. A failed control test should change a residual score within days, and a residual score above appetite should produce either a treatment or a signed acceptance.
What artifacts should a working program produce?
If you can hand these eight artifacts to a new auditor on day one, the program is real:
- A GRC charter naming decision rights, committees and escalation paths
- A policy set with owners, approval records and review dates
- A risk appetite statement with numeric thresholds
- A risk register with inherent and residual scores, owners and treatment dates
- A control library mapped to every framework in scope
- An evidence log showing when each control was last tested and by whom
- An exception register with approvers and expiry dates
- A board report template used every quarter
Most organizations already have half of these spread across documents, tickets and spreadsheets. The work is less about creating them than about linking them, so that a control failure visibly changes a risk score and a risk above appetite visibly lands on a board agenda.
A worked example: one control, three frameworks
Take one governance control and follow it through all three texts.
| Field | GOV-03: Annual review of the cyber risk management framework |
|---|---|
| Objective | Leadership confirms the cyber risk framework, appetite and policies still fit the threat picture and the business |
| Owner | CISO prepares; management body approves |
| Frequency | At least yearly, and after any major incident or significant audit finding |
| Evidence | Signed minutes; the reviewed appetite statement; a change log of policy updates |
| NIS2 | Art. 20(1) approval and oversight; Art. 21(2)(f) effectiveness assessment |
| DORA | Art. 6(5) review at least once a year and after major ICT-related incidents; Art. 5(2)(d) approval of strategy and risk tolerance |
| NIST CSF 2.0 | GV.OV-01 outcomes reviewed to adjust strategy; GV.RM-02 appetite statements maintained |
| Test | Minutes exist, are dated within 12 months, and record a decision on appetite |
One meeting, one set of minutes, three frameworks satisfied. Run the same exercise for incident response, supplier assessment and access control and you will find that most of your control set maps to more than one obligation.
That reuse is the main efficiency GRC offers a security team. Without it, the ISO 27001 auditor, the NIS2 assessment and the customer questionnaire each trigger a fresh round of evidence collection for the same control.
Where does third-party risk sit in cyber GRC?
Suppliers are where governance, risk and compliance meet most visibly. NIS2 Article 21(2)(d) lists supply chain security among the required measures. DORA Article 5(2)(h) makes the management body approve the policy on ICT third-party service providers. NIST CSF 2.0 places supply chain risk management (GV.SC) inside GOVERN, not in a separate procurement process.
In practice that means four things:
- One supplier register shared by procurement, security and compliance
- Criticality tiers that decide how deep each assessment goes
- Assessment findings that become risks in the main register, with owners
- Contract terms and review frequency that follow from the tier
When the supplier list and the risk register live in separate files, a critical vendor's failed assessment rarely reaches the people who could act on it. Linking the two is usually the fastest visible win a new GRC program can deliver.
How do you measure a cyber GRC program?
Measure decisions and coverage, not activity. A short scorecard works better than a long one.
| Area | Metric | How to calculate it |
|---|---|---|
| Governance | Policy currency | Policies within their review date ÷ total policies |
| Governance | Decisions on time | Risks above appetite with a leadership decision within the agreed period ÷ risks above appetite |
| Risk | Risks above appetite | Open risks whose residual score exceeds appetite, shown as a trend |
| Risk | Treatment on schedule | Treatment actions closed by their due date ÷ actions due |
| Compliance | Control test coverage | Controls tested in the current cycle ÷ controls in scope |
| Compliance | Control pass rate | Controls that passed their last test ÷ controls tested |
| Compliance | Expired exceptions | Exceptions past expiry and neither renewed nor closed |
| Third parties | Critical suppliers assessed | Critical suppliers with a current assessment ÷ critical suppliers |
Set the targets yourself and review them every quarter. A board needs three of these on one page, usually risks above appetite, decisions on time and control pass rate, not all eight.
Avoid metrics that only count effort, such as the number of policies written or questionnaires sent. They rise while risk stays flat, and leadership learns to ignore them.
What goes in a quarterly board report?
Board reports fail when they list activity. A report that supports the approval and oversight NIS2 Article 20(1) and DORA Article 5(2) expect from leadership fits on two pages:
- Risks above appetite, each with owner, residual score, trend and the decision requested.
- Changes since last quarter: new risks, closed risks, and scores that moved, with the reason.
- Control health: test coverage and pass rate for the cycle, naming the failed controls that matter.
- Exceptions close to expiry, and any accepted above the usual authority level.
- Third parties: critical suppliers without a current assessment, and supplier incidents.
- Incidents: significant incidents, whether reporting deadlines were met, and what changed as a result.
- Decisions needed: a short list to approve, defer or reject, with a recommendation for each.
The last section matters most. Minutes that record a decision on each item are the evidence that leadership approved and oversaw the measures, which is exactly what both regulations ask for.
How mature is your GRC program?
NIST CSF 2.0 describes four Tiers that characterize the rigor of an organization's cybersecurity risk governance and management practices: Partial (Tier 1), Risk Informed (Tier 2), Repeatable (Tier 3) and Adaptive (Tier 4). The observable signs below are our interpretation, not NIST's wording.
| Tier | What you would observe |
|---|---|
| Tier 1, Partial | Policies exist but are not enforced; risks are discussed but not recorded; compliance is a project before each audit |
| Tier 2, Risk Informed | A risk register exists and leadership sees it, but not on a fixed cycle; controls map to one framework |
| Tier 3, Repeatable | Appetite is approved; risks, controls and evidence are linked; tests follow a calendar; board reporting is regular |
| Tier 4, Adaptive | Risk scores move with live signals such as control failures, findings and supplier changes; strategy adjusts from the program's own results |
If you are approaching NIS2 or DORA for the first time, expect to start at Tier 1 or 2. Tier 3 is the realistic medium-term goal, because that is where approval, oversight and periodic review stop depending on individual effort.
What are the most common cyber GRC mistakes?
- Buying a tool before choosing a method. A platform will faithfully automate an undefined scoring scale.
- One framework at a time. Running ISO 27001, then NIS2, then DORA as separate projects multiplies evidence work.
- Appetite in words only. "Low appetite for cyber risk" cannot be compared with a score. Use a number.
- IT owning business risks. The CISO coordinates; the leader whose process is at risk owns the decision.
- Evidence by email. If evidence cannot be found without asking someone, it does not exist for an auditor.
- Reports that ask for nothing. A board report without a decision request gets no decision back.
Spreadsheets or a GRC platform?
Spreadsheets are a reasonable start. A short risk register and a control list in a shared workbook can carry a small organization through its first assessment.
They stop working when the links matter more than the lists. The signs are familiar: the same control is tested separately for ISO 27001 and NIS2; nobody can say which risks a failed control affects; evidence lives in email; the supplier list and the risk register disagree.
That is the point at which a platform earns its cost. ResilientX GRC software keeps frameworks, controls, risks, vendors, assets, policies, audits and evidence in one environment, with cross-mapped libraries for NIS2, DORA, ISO 27001 and GDPR, so a control assessed once is reused across every framework that shares the requirement. Auditors get a signed-invite portal that shows only in-scope controls, evidence and policies, backed by tamper-evident logs.
The risk register supports configurable 5x5, 4x4 or 3x3 scoring, appetite thresholds, and links from each risk to controls, findings, assets and vendors. Existing Excel registers can be imported instead of rebuilt. See ResilientX GRC in action if your registers have already outgrown the spreadsheet.
Cyber GRC build checklist
Use this as a first-year plan. Each line has a "done when" test so progress is binary. For a phased version with milestones, see the GRC implementation roadmap.
| # | Item | Done when |
|---|---|---|
| 1 | Scope | Applicable regulations, frameworks and business units are listed and approved |
| 2 | Charter | Decision rights, committees and escalation paths are signed off by leadership |
| 3 | Appetite | A risk appetite statement with numeric thresholds is approved by the management body |
| 4 | Scoring method | One likelihood and impact scale, with definitions, is used for every risk |
| 5 | Register | Top risks are recorded with owners, inherent and residual scores and treatment dates |
| 6 | Control library | Controls are written once and mapped to every framework in scope |
| 7 | Evidence | Each control has a defined evidence item and a test frequency |
| 8 | Exceptions | Every policy exception has an approver and an expiry date |
| 9 | Suppliers | Critical suppliers are tiered and have a current assessment |
| 10 | Incidents | The NIS2 24-hour and 72-hour reporting path is documented and rehearsed, if NIS2 applies |
| 11 | Board reporting | A quarterly report shows risks against appetite and the decisions needed |
| 12 | Annual review | Framework, appetite and policies are reviewed and the review is minuted |
Items 1 to 5 are the foundation and usually come first. Items 6 to 9 are where most of the effort goes. Items 10 to 12 turn a project into a program, because they repeat.
Frequently asked questions
Is GRC the same as cybersecurity?
No. Cybersecurity is the set of technical and operational controls that protect systems and data. GRC is the management layer around them: it decides which risks those controls should address, who can accept what remains, and how you prove the controls work. A strong security team without GRC is usually busy without being able to show the board whether risk is going down.
Which framework should a European company start with?
Start with the obligation that actually binds you: NIS2 if you are an essential or important entity, DORA if you are a financial entity in its scope. Use ISO/IEC 27001 or NIST CSF 2.0 as the structure of your control library, then map the legal requirements onto it. The crosswalk table above shows how closely the three already line up.
Do small companies need a cyber GRC program?
They need the decisions, not the bureaucracy. NIS2 Article 21(1) asks for measures that are appropriate and proportionate, which leaves room for a light program. For a small company that can mean one page of appetite, a short register of top risks with named owners, a mapped control list and a quarterly review. What it cannot mean is no owner.
How long does it take to stand up cyber GRC?
It depends on scope and on what already exists. A first useful version, meaning scope, appetite, a register of top risks and a mapped control list, can be in place well before a full program exists. A repeatable program takes at least one full annual cycle, because the annual review, the test calendar and quarterly board reporting are part of what makes it repeatable.
What is the difference between GRC and ERM?
Enterprise risk management covers every category of risk, from credit and market to operational and strategic. Cyber GRC is narrower in subject and wider in mechanics: it covers cyber and ICT risk plus the policies, controls and compliance evidence around it. NIST CSF 2.0 subcategory GV.RM-03 expects cybersecurity risk activities to feed enterprise risk management. More in our GRC risk management guide.
The short version
GRC in cybersecurity comes down to three habits: decisions made by people with the authority to make them, risks ranked by what they threaten, and controls that can prove they work. NIS2 and DORA have turned those habits into legal duties for leadership. Build the loop once, map it to every framework you face, and review it every year.
See ResilientX GRC in action — book a demo and see your NIS2, DORA and ISO 27001 controls in one connected record.