Who it applies to
NIS2 covers essential and important entities. Under Article 3(1), entities of a type listed in Annex I (sectors of high criticality, such as energy, transport, banking, health and digital infrastructure) that exceed the ceilings for medium-sized enterprises are essential, together with certain other categories named in that paragraph. Under Article 3(2), Annex I or Annex II entities that do not qualify as essential are important. Annex II covers other critical sectors, such as postal and courier services, waste management, the manufacture, production and distribution of chemicals, the production, processing and distribution of food, and listed manufacturing subsectors. Article 2(1) applies the directive to entities of a type in Annex I or II that qualify as medium-sized enterprises or exceed those ceilings and that provide their services or carry out their activities within the Union, and Article 2(2) applies it regardless of size to certain listed categories.
The size test is the one in Article 2(1), which refers to Article 2 of the Annex to Commission Recommendation 2003/361/EC. Article 2(1) also states that Article 3(4) of that Annex does not apply for the purposes of the directive. The definitions should be checked against the Recommendation for the specific entity.
A supplier that is not itself in scope is not bound by the directive as such. Its customers that are in scope must address the relationship under Article 21(2)(d), usually through contract terms. For certain digital-infrastructure and ICT-service entities, Commission Implementing Regulation (EU) 2024/2690 of 17 October 2024 lays down the technical and methodological requirements of the measures in Article 21(2) and further specifies when an incident is considered significant. Under its Article 1, it covers DNS service providers, TLD name registries, cloud computing service providers, data centre service providers, content delivery network providers, managed service providers, managed security service providers, providers of online market places, of online search engines and of social networking services platforms, and trust service providers. For other essential and important entities, Article 21(5) of the directive allows, but does not require, the Commission to adopt further implementing acts.
Sources: Directive (EU) 2022/2555, Articles 2 and 3, Annexes I and II (Official Journal); Implementing Regulation (EU) 2024/2690, Article 1 (Official Journal)
Key dates
| Date | What happens | Source |
|---|---|---|
| Directive published in the Official Journal (OJ L 333) | Directive (EU) 2022/2555 on EUR-Lex | |
| Entry into force, on the twentieth day after publication (Article 45) | Directive (EU) 2022/2555, Article 45 on EUR-Lex | |
| Transposition deadline: member states adopt and publish their national measures (Article 41(1)) | Directive (EU) 2022/2555, Article 41 on EUR-Lex | |
| Member states apply their national measures from this date (Article 41(1)) | Directive (EU) 2022/2555, Article 41 on EUR-Lex | |
| Member states establish lists of essential and important entities (Article 3(3)) | Directive (EU) 2022/2555, Article 3 on EUR-Lex | |
| First Commission review of the directive, then every 36 months (Article 40(1)) | Directive (EU) 2022/2555, Article 40 on EUR-Lex |
What NIS2 requires from a security team in practice
NIS2 is a directive. It sets goals and minimum content, and each EU member state turns it into national law, which can add detail, deadlines and supervisory practice. The transposition deadline was 17 October 2024, and member states were to apply their measures from 18 October 2024 (Article 41(1)). The national law of each country has to be checked locally.
For a security team, three parts of the text carry most of the work. Article 21 requires “appropriate and proportionate technical, operational and organisational measures” based on an all-hazards approach, including at least ten listed areas. Article 20 makes the management body approve those measures and oversee them. Article 23 sets a timed process for reporting significant incidents. The sections below take each in turn.
Turning Article 21 into a work plan
Article 21(2) lists ten areas that the measures must at least include, from (a) policies on risk analysis and information system security to (j) multi-factor or continuous authentication and secured communications, where appropriate. The list is not a product checklist. It asks for policies and procedures, applied in practice, with a named owner and evidence that they work.
A workable approach is to take the ten points as rows in a register. For each row, record the current policy, the owner, the evidence you could show an auditor today, and the gap. Article 21(4) adds a duty that applies once gaps appear: an entity that finds it does not comply must take, without undue delay, all necessary, appropriate and proportionate corrective measures. A register that lists gaps without dates or owners shows no progress on that duty.
Where security testing fits, and where it does not
The directive does not name penetration testing. It does name two areas that test results can feed. Article 21(2)(e) covers security in acquisition, development and maintenance, including vulnerability handling and disclosure. Article 21(2)(f) requires policies and procedures to assess the effectiveness of cybersecurity risk-management measures. A test does not write those policies. It can provide dated, reproducible evidence about the systems it reached.
AssurePort’s Web Pentest ($399) and API Pentest ($399) run authorised tests of web applications and APIs. Every scan needs either a verified asset or a declaration of authorisation from the customer. Each confirmed finding carries a reproducible proof, and anything the engine cannot prove is labelled unconfirmed rather than counted. The GitHub SAST engine covers secrets, dependencies and CI/CD configuration in repositories, which is relevant to Article 21(2)(d) when dependencies are the question. Scans are paid per engine.
A penetration test does not show governance, staff training, business continuity and backup restoration, the cryptography policy, HR security, or MFA rollout across the organisation. Points (b), (c), (g) and the policy side of (h), (i) and (j) need their own evidence. A test is one part of the evidence file, not the whole of it.
Incident reporting under Article 23: the timeline
Under Article 23(1), essential and important entities must notify, without undue delay, their CSIRT (computer security incident response team) or competent authority of any incident that has a significant impact on the provision of their services. Article 23(3) defines a significant incident as one that has caused or is capable of causing severe operational disruption of the services or financial loss for the entity, or that has affected or is capable of affecting other natural or legal persons by causing considerable material or non-material damage.
The clock in Article 23(4) starts when the entity becomes aware of the significant incident. The early warning is due within 24 hours and, where applicable, says whether the incident is suspected of being caused by unlawful or malicious acts or could have a cross-border impact. The incident notification is due within 72 hours and gives an initial assessment of severity and impact and, where available, indicators of compromise. An intermediate report is due when a CSIRT or the competent authority asks for one. The final report is due not later than one month after the submission of the incident notification under point (b). If the incident is still ongoing at that time, a progress report is due instead, with a final report within one month of the handling of the incident (Article 23(4)(e)). The table below sets out each stage.
Two points deserve attention. The clock runs from awareness, not from the end of an investigation, so the playbook should decide in advance who can determine significance. And the directive sets the outer limits, while national law can add procedural steps, so the local reporting route must be written into the playbook as well.
Management accountability and penalties
Under Article 20(1), management bodies of essential and important entities must approve the cybersecurity risk-management measures, oversee their implementation, and can be held liable for infringements of Article 21. Under Article 20(2), members of management bodies must follow training, and entities are encouraged to offer similar training to employees.
Fines are set by Article 34. For infringements of Article 21 or 23, essential entities are subject to fines with a maximum of at least EUR 10 000 000 or at least 2 % of the total worldwide annual turnover in the preceding financial year of the undertaking to which the entity belongs, whichever is higher (Article 34(4)). For important entities the figures are at least EUR 7 000 000 or 1.4 %, again whichever is higher (Article 34(5)). “At least” means the directive sets a floor for the maximum that national law must provide. Actual figures depend on national transposition.
The supervisory tools also differ. Essential entities can be subject to on-site inspections, security audits by an independent body or the authority, security scans, and requests for evidence such as the results of audits by a qualified auditor (Article 32(2)). If enforcement measures under Article 32(4), points (a) to (d) and (f), are ineffective and the entity does not act by a deadline set by the authority, the authority can suspend temporarily, or request the suspension of, a certification or authorisation, and can request that the relevant bodies, courts or tribunals temporarily prohibit a person discharging managerial responsibilities at chief executive officer or legal representative level from exercising managerial functions (Article 32(5)). For important entities, authorities act through ex post supervisory measures when provided with evidence, indication or information that the entity allegedly does not comply (Article 33(1)), and the enforcement powers in Article 33(4) do not include a suspension or a management ban. The directive does not say whether a particular test report counts as a security audit under Article 32 or 33.
Starting this quarter
Start with scope: decide whether the entity is essential or important under Articles 2 and 3 and national law, and find out what your national authority requires you to submit (Article 3(4)). Next, build the Article 21 register and assign owners. Then run an Article 23 exercise, a tabletop that starts from awareness of a significant incident and walks through the early warning, the 72-hour notification and the final report, with the contact route to your CSIRT or authority written down.
Put evidence on a schedule. Dated test reports, restore tests, access reviews and training records should sit in one place, so the evidence for each Article 21 point can be produced when asked. The free security headers check, email authentication check and DNSSEC check are passive, one-off reads of public configuration. They are not penetration tests. For the wider picture, see our ISO 27001 Annex A matrix and how penetration testing is scoped.
Obligations and where testing can contribute evidence
| Reference | What the text requires | Where a penetration test can contribute evidence |
|---|---|---|
| Art. 20(1)–(2) | Management body approval, oversight and training Management bodies must approve the cybersecurity risk-management measures and oversee their implementation, and they can be held liable for infringements of Article 21. Members must follow training, and entities are encouraged to offer similar training to employees. Source: Directive (EU) 2022/2555, Art. 20 (EUR-Lex) | A dated test report can be among the material a management body reviews, but approval, oversight and training records are not produced by a test. |
| Art. 21(2)(a) | Risk analysis and information system security policies Policies on risk analysis and on the security of network and information systems. Source: Directive (EU) 2022/2555, Art. 21(2)(a) (EUR-Lex) | Test findings can feed a risk analysis by showing which weaknesses are reproducible on the systems tested. They do not replace the analysis or the policy. |
| Art. 21(2)(b) | Incident handling Measures for incident handling, one of the listed areas that sits alongside the reporting duties in Article 23. Source: Directive (EU) 2022/2555, Art. 21(2)(b) (EUR-Lex) | Outside what a test can evidence. |
| Art. 21(2)(c) | Business continuity, backup management, disaster recovery and crisis management Business continuity measures, such as backup management and disaster recovery, and crisis management. Source: Directive (EU) 2022/2555, Art. 21(2)(c) (EUR-Lex) | Outside what a test can evidence. |
| Art. 21(2)(d) | Supply chain security Supply chain security, including security-related aspects of the relationships between the entity and its direct suppliers or service providers. Under Article 21(3), entities take into account the vulnerabilities specific to each direct supplier and service provider and the overall quality of their products and cybersecurity practices, including their secure development procedures. Source: Directive (EU) 2022/2555, Art. 21(2)(d) and 21(3) (EUR-Lex) | A supplier’s shared test report can be one input to a supplier assessment. A test of your own systems does not show how a supplier secures its own. |
| Art. 21(2)(e) | Security in acquisition, development and maintenance, including vulnerability handling and disclosure Security in the acquisition, development and maintenance of network and information systems, including vulnerability handling and disclosure. Source: Directive (EU) 2022/2555, Art. 21(2)(e) (EUR-Lex) | Tests give dated evidence of which application and API weaknesses were found and whether they can be exploited. A vulnerability disclosure policy and the handling process remain the entity’s own documents. |
| Art. 21(2)(f) | Assessing the effectiveness of measures Policies and procedures to assess whether the cybersecurity risk-management measures work. Source: Directive (EU) 2022/2555, Art. 21(2)(f) (EUR-Lex) | A penetration test is one way to check that technical measures work, and its results can feed that assessment. The assessment policy itself is not a test output. |
| Art. 21(2)(g) | Basic cyber hygiene and cybersecurity training Basic cyber hygiene practices and cybersecurity training. Source: Directive (EU) 2022/2555, Art. 21(2)(g) (EUR-Lex) | Outside what a test can evidence. |
| Art. 21(2)(h) | Cryptography and encryption policies Policies and procedures regarding the use of cryptography and, where appropriate, encryption. Source: Directive (EU) 2022/2555, Art. 21(2)(h) (EUR-Lex) | Web, API and Network / Host tests report weak transport encryption on the endpoints they reach. The cryptography policy and key management are outside what a test covers. |
| Art. 21(2)(i) | HR security, access control and asset management Human resources security, access control policies and asset management. Source: Directive (EU) 2022/2555, Art. 21(2)(i) (EUR-Lex) | Tests can show access-control weaknesses on the tested applications and directory services, such as broken authorisation or misconfigured delegation. Joiner and leaver processes and asset inventories are not tested. |
| Art. 21(2)(j) | Multi-factor authentication and secured communications The use of multi-factor authentication or continuous authentication solutions, secured voice, video and text communications and secured emergency communication systems within the entity, where appropriate. Source: Directive (EU) 2022/2555, Art. 21(2)(j) (EUR-Lex) | A test can show whether tested login paths can be reached without the second factor. It does not show MFA rollout across the organisation. |
| Art. 23(1)–(4) | Significant incident notification Significant incidents must be notified without undue delay to the CSIRT or competent authority: an early warning within 24 hours, an incident notification within 72 hours, an intermediate report on request, and a final report not later than one month after the submission of the incident notification. Source: Directive (EU) 2022/2555, Art. 23 (EUR-Lex) | Outside what a test can evidence. |
Reporting deadlines
| Deadline | What must be reported | To | Reference |
|---|---|---|---|
| Without undue delay and in any event within 24 hours of becoming aware of the significant incident | Early warning, indicating, where applicable, whether the incident is suspected of being caused by unlawful or malicious acts or could have a cross-border impact | The CSIRT or, where applicable, the competent authority | Art. 23(4)(a) |
| Without undue delay and in any event within 72 hours of becoming aware of the significant incident | Incident notification, updating the early warning where applicable and giving an initial assessment of severity and impact, and indicators of compromise where available | The CSIRT or, where applicable, the competent authority | Art. 23(4)(b) |
| On request of the CSIRT or the competent authority | Intermediate report on relevant status updates | The CSIRT or, where applicable, the competent authority | Art. 23(4)(c) |
| Not later than one month after the submission of the incident notification under point (b) | Final report: detailed description of the incident including severity and impact, the type of threat or root cause likely to have triggered it, applied and ongoing mitigation measures, and any cross-border impact where applicable | The CSIRT or, where applicable, the competent authority | Art. 23(4)(d) |
| Without undue delay, where appropriate | Notice to service recipients potentially affected by a significant cyber threat, with any measures or remedies they can take, and the threat itself where appropriate | Service recipients potentially affected | Art. 23(2) |
Frequently asked questions
Does NIS2 require a penetration test?
No specific test is named. Article 21(2)(e) asks for security in acquisition, development and maintenance, including vulnerability handling and disclosure, and Article 21(2)(f) asks for policies and procedures to assess the effectiveness of the measures. A penetration test can support both as dated, reproducible evidence, but it is one input among several and does not replace the policies themselves.
How fast must a significant incident be reported under NIS2?
Under Article 23(4)(a), an early warning is due within 24 hours of becoming aware of the significant incident. The incident notification follows within 72 hours (Article 23(4)(b)). An intermediate report is due when the CSIRT or competent authority asks for one (Article 23(4)(c)), and a final report is due not later than one month after the submission of the incident notification under point (b) (Article 23(4)(d)). National law may add steps.
Does a penetration test show that an organisation meets NIS2?
No. The directive covers governance, training, incident reporting, business continuity and supplier relationships as well as technical measures. A test can support evidence for some of these points, such as Article 21(2)(e), but no single test result establishes that an entity meets the directive.
How do we know whether we are an essential or an important entity?
Articles 3(1) and 3(2) decide the status from the sector annex, the size test in Article 2(1), and the categories listed in Article 3(1). Member States had to establish lists of essential and important entities by 17 April 2025 (Article 3(3)). Check your national law and your authority’s guidance to confirm your status.
What fines apply for NIS2 infringements?
For infringements of Article 21 or 23, Article 34(4) sets maximum fines for essential entities of at least EUR 10 million or 2 % of the total worldwide annual turnover in the preceding financial year of the undertaking, whichever is higher. Article 34(5) sets at least EUR 7 million or 1.4 % for important entities. For essential entities, if certain earlier enforcement measures are ineffective and a deadline is missed, authorities can also request a temporary prohibition on a person at chief executive officer or legal representative level exercising managerial functions (Article 32(5)). The actual amounts depend on national law.
Related
Testing is one input to your Article 21 work
Web Pentest ($399), API Pentest ($399) and Network / Host Pentest ($579) are paid per scan. Each confirmed finding carries a reproducible proof, and unproven findings are labelled unconfirmed.