Key takeaways
Understand Belgium's NIS2 approach, the CCB, CyFun, governance, risk management, incidents and evidence.
A relevant audit starts with the services the business must continue to deliver, the information it cannot afford to lose and the commitments it must meet. Technical scope follows. This prevents teams from spending time on a peripheral device while overlooking a cloud account, supplier or restoration process on which operations genuinely depend.
The NIST Cybersecurity Framework 2.0 organises risk management around six functions — govern, identify, protect, detect, respond and recover — while the Centre for Cybersecurity Belgium's CyberFundamentals Framework translates these principles into practical, proportionate measures. Frameworks do not replace judgement; they reduce blind spots and create a common management and technical language. [1][2]
In this guide, a control is reliable only when it has an owner, consistent configuration, known coverage and recent evidence. A purchased licence, enabled checkbox or old document is not enough. Effectiveness is demonstrated through use, monitoring, testing and the ability to remediate exceptions.
The result must remain understandable over time. For each conclusion, record the observation date, covered systems and populations, method, sample and limitations. Separate observed facts from statements and auditor assumptions. This traceability lets another person understand the rating, repeat the verification after remediation and measure progress during the next review. It also prevents two common errors: assuming a control applies everywhere when only part of the scope was checked, or keeping an action open indefinitely without an objective closure criterion. Provide two reporting levels: a concise management view for decisions and technical detail for implementation. Both should use the same priorities and clearly identify risks that are accepted, transferred, reduced or avoided. The report then becomes a governed working instrument rather than a snapshot forgotten after presentation.
- ✓NIS2, CyFun and obligations
Compliance starts by establishing applicability, then linking each obligation to an owner, control and evidence. A matrix without implementation creates false assurance.
- ✓Governance, accountability and risk
Verify who decides, performs, checks and accepts risk. Informal accountability delays incident response and makes exceptions difficult to govern.
- ✓Suppliers and the supply chain
Providers may administer systems, host data or support critical activities. Their controls, contracts and response capabilities affect your risk.
- ✓Logging, detection and incident response
Without useful logs and clear roles, compromise can remain invisible. Collection should support concrete detection and response scenarios.
- ✓Continuity, recovery and crisis management
Continuity coordinates business priorities, human and technical dependencies and decisions under pressure. An untested plan may be unusable.
NIS2
NIS2, CyFun and obligations
Compliance starts by establishing applicability, then linking each obligation to an owner, control and evidence. A matrix without implementation creates false assurance. [1]
This domain must be examined through real operations rather than a declared procedure alone. The auditor links people, technology and data flows to the business scenario in this guide. Exceptions matter too: a protection enabled broadly may still leave critical accounts, devices or data outside its coverage.
The method compares the expected, configured and actually observed states. Interviews explain intent; administration screens, exports, logs and tests demonstrate application. Where intrusive verification is not authorised, that limitation must be explicit in the report to avoid an overconfident conclusion.
Checks to perform
- Confirm legal scope and affected entities
- Determine the appropriate CyFun level
- Connect risk measures, incidents and governance
- Check management training and involvement
- Organise evidence and follow-up dates
Evidence to request
- Scope assessment and decisions
- Requirement-control-evidence matrix
- Incident and training records
Each gap is then qualified by likelihood, impact and ease of exploitation. The recommendation should name the expected result, action owner, dependencies and evidence needed for closure. This discipline turns a technical list into a governance decision.
Expected decisions and actions
- 1Validate applicability
- 2Close priority gaps
- 3Prepare demonstrability
NIS2
Governance, accountability and risk
Verify who decides, performs, checks and accepts risk. Informal accountability delays incident response and makes exceptions difficult to govern. [2]
This domain must be examined through real operations rather than a declared procedure alone. The auditor links people, technology and data flows to the business scenario in this guide. Exceptions matter too: a protection enabled broadly may still leave critical accounts, devices or data outside its coverage.
The method compares the expected, configured and actually observed states. Interviews explain intent; administration screens, exports, logs and tests demonstrate application. Where intrusive verification is not authorised, that limitation must be explicit in the report to avoid an overconfident conclusion.
Checks to perform
- Name an executive sponsor and operational owner
- Link IT risks to the enterprise risk register
- Check policy approval and review dates
- Identify risk acceptances and expiry dates
- Separate execution from validation
Evidence to request
- Responsibility matrix
- Risk register and minutes
- Approved policies and metrics
Each gap is then qualified by likelihood, impact and ease of exploitation. The recommendation should name the expected result, action owner, dependencies and evidence needed for closure. This discipline turns a technical list into a governance decision.
Expected decisions and actions
- 1Assign every major risk
- 2Set a governance rhythm
- 3Document exceptions
NIS2
Suppliers and the supply chain
Providers may administer systems, host data or support critical activities. Their controls, contracts and response capabilities affect your risk. [3]
This domain must be examined through real operations rather than a declared procedure alone. The auditor links people, technology and data flows to the business scenario in this guide. Exceptions matter too: a protection enabled broadly may still leave critical accounts, devices or data outside its coverage.
The method compares the expected, configured and actually observed states. Interviews explain intent; administration screens, exports, logs and tests demonstrate application. Where intrusive verification is not authorised, that limitation must be explicit in the report to avoid an overconfident conclusion.
Checks to perform
- Inventory critical suppliers and access
- Review security, notification and exit clauses
- Check third-party accounts and remote access
- Assess concentration and dependencies
- Plan exit, data return and continuity
Evidence to request
- Supplier register and contracts
- Available assurance reports
- Third-party account list
Each gap is then qualified by likelihood, impact and ease of exploitation. The recommendation should name the expected result, action owner, dependencies and evidence needed for closure. This discipline turns a technical list into a governance decision.
Expected decisions and actions
- 1Classify third parties
- 2Reduce standing access
- 3Formalise notification and exit
NIS2
Logging, detection and incident response
Without useful logs and clear roles, compromise can remain invisible. Collection should support concrete detection and response scenarios. [4]
This domain must be examined through real operations rather than a declared procedure alone. The auditor links people, technology and data flows to the business scenario in this guide. Exceptions matter too: a protection enabled broadly may still leave critical accounts, devices or data outside its coverage.
The method compares the expected, configured and actually observed states. Interviews explain intent; administration screens, exports, logs and tests demonstrate application. Where intrusive verification is not authorised, that limitation must be explicit in the report to avoid an overconfident conclusion.
Checks to perform
- Identify critical log sources
- Check alerts, recipients and coverage hours
- Test escalation with a realistic scenario
- Check retention, time and access
- Connect technical, legal and communication response
Evidence to request
- Log catalogue and alert rules
- Incident tickets and reports
- Response plan and contacts
Each gap is then qualified by likelihood, impact and ease of exploitation. The recommendation should name the expected result, action owner, dependencies and evidence needed for closure. This discipline turns a technical list into a governance decision.
Expected decisions and actions
- 1Cover critical signals
- 2Define escalation
- 3Run an exercise
NIS2
Continuity, recovery and crisis management
Continuity coordinates business priorities, human and technical dependencies and decisions under pressure. An untested plan may be unusable. [5]
This domain must be examined through real operations rather than a declared procedure alone. The auditor links people, technology and data flows to the business scenario in this guide. Exceptions matter too: a protection enabled broadly may still leave critical accounts, devices or data outside its coverage.
The method compares the expected, configured and actually observed states. Interviews explain intent; administration screens, exports, logs and tests demonstrate application. Where intrusive verification is not authorised, that limitation must be explicit in the report to avoid an overconfident conclusion.
Checks to perform
- Identify critical activities, RTO and RPO
- Document dependencies, contacts and workarounds
- Align IT recovery and business continuity
- Plan internal, customer and authority communication
- Run an exercise and track lessons
Evidence to request
- Business impact analysis
- Continuity and recovery plans
- Exercise reports
Each gap is then qualified by likelihood, impact and ease of exploitation. The recommendation should name the expected result, action owner, dependencies and evidence needed for closure. This discipline turns a technical list into a governance decision.
Expected decisions and actions
- 1Validate priorities
- 2Test procedures
- 3Remove single points of failure
Turn the audit into a roadmap
Turn the audit into a roadmap
A good audit does not end with a score. It organises the move from the observed state to a chosen level of control, with an owner, deadline and closure evidence for every action.
Frame the risk
Confirm critical activities, important data, dependencies, obligations and risk appetite with management. Without that context, every fix appears urgent and budgets become scattered.
Treat immediate exposure
Start with scenarios combining high likelihood and impact: weakly protected administrator accounts, backups reachable from production, exposed or unsupported systems and insufficient detection capability.
Stabilise the foundations
Assign responsibilities, complete the inventory, standardise configurations, document procedures and introduce recurring checks. Security should be repeatable rather than dependent on one person.
Prove and steer
Retain configuration reports, logs, test records and risk acceptance decisions. A small number of stable indicators is more useful than a crowded dashboard that triggers no decisions.
Reassess
Schedule a review after major changes and at least at a defined frequency. An audit is a snapshot; new users, cloud services, suppliers and equipment continuously change risk.
Preparation checklist
Preparation checklist
Collect these items before the interview. Missing evidence does not automatically mean the control is absent, but it increases verification time and weakens assurance.
- Confirm legal scope and affected entities
- Determine the appropriate CyFun level
- Scope assessment and decisions
- Name an executive sponsor and operational owner
- Link IT risks to the enterprise risk register
- Responsibility matrix
- Inventory critical suppliers and access
- Review security, notification and exit clauses
- Supplier register and contracts
- Identify critical log sources
- Check alerts, recipients and coverage hours
- Log catalogue and alert rules
Frequently asked questions
Frequently asked questions
How much time should this type of audit take?+
It depends on the number of sites, users, devices, cloud tenants and suppliers. A targeted review may take a few hours; a documented engagement covering governance, configurations and evidence usually requires several interviews and analysis time. Write down the scope before work begins.
Is a questionnaire enough to reach a conclusion?+
No. A questionnaire is excellent for directing discussion and revealing unknown areas. A professional conclusion still requires suitable evidence, observation or testing. Unverified answers must be clearly identified in the report.
Must every observation be fixed immediately?+
No. Actions should be prioritised by risk, business impact, effort and dependencies. Some measures reduce several risks at once; others can wait or be replaced by a documented compensating control.
Does an audit guarantee that no incident will occur?+
No audit can guarantee the absence of outages, errors or cyberattacks. It reduces uncertainty, highlights known exposure and improves prevention, detection, response and recovery. Its limitations should be explicit.
How often should the situation be reassessed?+
Set a risk-based frequency and add reviews after major changes: cloud migration, acquisition, new site, supplier change, incident, new obligation or business transformation. Critical actions deserve more frequent follow-up than the full audit.
Official sources and references
Official sources and references
The sources below support the principles and requirements discussed. GVISION recommendations are an operational interpretation for a business context and do not replace legal advice or official certification.
- 01Centre for Cybersecurity Belgium — NIS2 en Belgique2026https://ccb.belgium.be/fr/reglementation/nis2 ↗
- 02Union européenne — Directive (UE) 2022/2555 — NIS22022https://eur-lex.europa.eu/eli/dir/2022/2555/oj ↗
- 03Centre for Cybersecurity Belgium — CyFun® 2025 is here!2025https://ccb.belgium.be/news/cyfunr-2025-here ↗
- 04CCB · Safeonweb@work — CyberFundamentals Framework2025https://atwork.safeonweb.be/tools-resources/cyberfundamentals-framework ↗
- 05NIST — The NIST Cybersecurity Framework (CSF) 2.02024https://www.nist.gov/publications/nist-cybersecurity-framework-csf-20 ↗