NIS2 Significant Incident: What Actually Triggers the 24-Hour Clock

A ransomware strain encrypts a file server at a 60-person logistics client on Friday at 16:40. Backups restore it by 19:00. Nobody asks the question that decides the weekend: was that a NIS2 significant incident, and did the 24-hour clock just start?
Most MSPs know the 24/72-hour/one-month reporting sequence. Far fewer can say what flips an ordinary incident into a reportable one. The answer sits in one paragraph of Article 23, and getting it wrong costs you in both directions: over-reporting buries the regulator and your client in noise, under-reporting is a supervisory finding.
The NIS2 significant incident test in Article 23(3)
Directive (EU) 2022/2555 does not require you to report everything. Article 23(1) obliges essential and important entities to notify "any incident that has a significant impact on the provision of their services". Article 23(3) then defines when that is the case. An incident is significant if:
- it has caused or is capable of causing severe operational disruption of the services or financial loss for the entity concerned; or
- it has affected or is capable of affecting other natural or legal persons by causing considerable material or non-material damage.
Two words do the work: "capable of". You do not need the damage to have happened. A credible path to severe disruption is enough. That is why the Friday-night ransomware case is a judgement call about what could have happened, not only about the 19:00 restore.
Note also what the test is not. It is not a count of affected machines, and it is not the same as a GDPR personal data breach. Our guide to running both clocks on one breach covers the overlap.
Where the hard numbers come from: CIR 2024/2690
For most sectors, Article 23(3) stays qualitative at EU level and member states fill the gap. For one group, the Commission has written thresholds. Article 23(11) required implementing acts for DNS providers, TLD registries, cloud and data centre providers, CDNs, managed service providers, managed security service providers, online marketplaces, search engines and social networks. That became Implementing Regulation (EU) 2024/2690.
If your clients, or you, are an MSP or MSSP, this is your rulebook. Article 3(1) makes an incident significant where one or more of these is met:
- direct financial loss that exceeds EUR 500,000 or 5% of annual turnover in the preceding financial year, whichever is lower;
- exfiltration of trade secrets;
- the death of a natural person, or considerable damage to a natural person's health;
- successful, suspectedly malicious and unauthorised access to network and information systems that is capable of causing severe operational disruption;
- the recurring-incident rule in Article 4;
- the sector-specific criteria in Articles 5 to 14.
Note the "whichever is lower". For a small client with EUR 2 million turnover, 5% is EUR 100,000, so the threshold is far below half a million. Scheduled maintenance interruptions do not count.
Article 4 closes a loophole. Incidents that are individually below the line count together as one significant incident if they occurred at least twice within six months, share the same apparent root cause, and collectively meet the financial criterion. Three small phishing-led intrusions through the same unpatched gateway can add up to a reportable event.
Sector criteria: the 30-minute example
Articles 5 to 14 add technical triggers. For a DNS service provider, a recursive or authoritative resolution service that is completely unavailable for more than 30 minutes is significant, and so is an average response time above 10 seconds for more than an hour. We explain the full set in the 30-minute incident rule for digital infrastructure.
The point for a consultant: a sector rule can make an incident reportable that would not pass the general financial test. Check the sector article first, then the general criteria.
Everyone else: national thresholds fill the gap
For entities outside that list, Article 23(3) is the EU rule and national law supplies detail. Germany defines the term in the BSI Act, the Netherlands, Italy and Poland leave thresholds to secondary rules or regulators, and France and Spain have no national law yet. The result is that the same incident can be reportable in one country and a judgement call in another. Our transposition status overview shows who has a law.
NIS2 transposition by country
- AustriaIn force
- In force since October 1, 2026
- Act: NISG 2026 (BGBl. I Nr. 94/2025)
- Registration due December 31, 2026
Authority: Federal Office for Cybersecurity · Source
- NetherlandsIn force
- In force since August 15, 2026
- Act: Cyberbeveiligingswet (Stb. 2026, 187)
- Registration required, no transition period
Authority: NCSC (MijnNCSC) · Source
- LuxembourgIn force
- In force since May 10, 2026
- Act: Law of 5 May 2026
- Registration due July 10, 2026
Authority: ILR (CSSF for finance) · Source
- PolandIn force
- In force since April 3, 2026
- Act: UKSC (Dz.U. 2026 poz. 252)
- Registration due October 3, 2026
- Measures due April 3, 2027
Authority: Minister Cyfryzacji (Wykaz KSC) · Source
- PortugalIn force
- In force since April 3, 2026
- Act: Decree-Law 125/2025 (RJC)
Authority: CNCS · Source
- BulgariaIn force
- In force since February 13, 2026
- Act: Cybersecurity Act amendment (State Gazette No 17/2026)
Authority: Ministry of e-Government · Source
- MaltaIn force
- In force since January 23, 2026
- Act: S.L. 460.41
Authority: Critical Infrastructure Protection Department · Source
- SwedenIn force
- In force since January 15, 2026
- Act: SFS 2025:1506
Authority: NCSC-SE · Source
- EstoniaIn force
- In force since January 1, 2026
- Act: Cybersecurity Act (KüTS) amendment
- Registration within 3 months
Authority: RIA · Source
- GermanyIn force
- In force since December 6, 2025
- Act: NIS2UmsuCG
- Registration due March 6, 2026
- 17,729 registered by June 30, 2026
Authority: BSI · Source
- CzechiaIn force
- In force since November 1, 2025
- Act: Act No 264/2025 Coll.
- Registration within 60 days
Authority: NÚKIB · Source
- DenmarkIn force
- In force since July 1, 2025
- Act: NIS 2 Act No 434 of 6 May 2025
- Registration due October 1, 2025
Authority: Danish Agency for Societal Security · Source
- SloveniaIn force
- In force since June 19, 2025
- Act: ZInfV-1 (OG 40/25)
- Registration within 30 days
Authority: URSIV · Source
- CyprusIn force
- In force since April 25, 2025
- Act: Law 60(I)/2025
Authority: Digital Security Authority · Source
- FinlandIn force
- In force since April 8, 2025
- Act: Cybersecurity Act 124/2025
- Registration due May 8, 2025
- Measures due July 8, 2025
Authority: Traficom (NCSC-FI) · Source
- HungaryIn force
- In force since January 1, 2025
- Act: Act LXIX of 2024
Authority: SZTFH · Source
- SlovakiaIn force
- In force since January 1, 2025
- Act: Act No 366/2024 Coll.
- Registration within 60 days
Authority: NBÚ · Source
- RomaniaIn force
- In force since December 31, 2024
- Act: GEO 155/2024 (Law 124/2025)
- Registration within 30 days
Authority: DNSC · Source
- GreeceIn force
- In force since November 27, 2024
- Act: Law 5160/2024 (Gazette A' 195)
Authority: National Cybersecurity Authority · Source
- BelgiumIn force
- In force since October 18, 2024
- Registration due March 18, 2025
Authority: CCB · Source
- LithuaniaIn force
- In force since October 18, 2024
- Act: Law No XIV-2902
Authority: NCSC (MoND) · Source
- ItalyIn force
- In force since October 16, 2024
- Act: D.Lgs. 138/2024
- Baseline measures due by October 2026 (entities listed in 2025)
Authority: ACN · Source
- LatviaIn force
- In force since September 1, 2024
- Act: National Cybersecurity Law
Authority: National Cybersecurity Centre · Source
- CroatiaIn force
- In force since February 15, 2024
- Act: Cybersecurity Act (OG 14/2024)
Authority: NCSC (SOA) · Source
- FranceNot yet transposed
- Senate adopted the law on March 12, 2025
- Regulator published its framework on March 17, 2026
- Referred to the Court of Justice on July 8, 2026
- Pre-registration open
Authority: ANSSI · Source
- SpainNot yet transposed
- Draft bill approved on January 14, 2025
- Referred to the Court of Justice on July 8, 2026
- The NIS1 regime still applies
Authority: CCN / INCIBE · Source
- IrelandNot yet transposed
- Referred to the Court of Justice on July 8, 2026
- Bill not yet enacted
Authority: NCSC Ireland · Source
Status as of September 28, 2026. Other member states: see the country-by-country timeline.
What starts the clock
Once an incident is significant, Article 23(4) sets the sequence: an early warning within 24 hours of becoming aware, an incident notification within 72 hours, intermediate reports on request, and a final report within one month of the 72-hour notification. The trigger is awareness that the incident is significant, not the moment the first alert fires. But that is no licence to delay classification. A regulator will look at when a reasonable team would have known.
NIS2 Incident Reporting Timeline
Early Warning
Notify the competent authority (CSIRT/NCA) within 24 hours of becoming aware of a significant incident.
Incident Notification
Submit a detailed notification within 72 hours with an initial assessment of severity, impact and indicators of compromise.
Final Report
Deliver a comprehensive final report within one month covering root cause, remediation taken and cross-border impact.
Early Warning
Notify the competent authority (CSIRT/NCA) within 24 hours of becoming aware of a significant incident.
Incident Notification
Submit a detailed notification within 72 hours with an initial assessment of severity, impact and indicators of compromise.
Final Report
Deliver a comprehensive final report within one month covering root cause, remediation taken and cross-border impact.
The early warning also has to say whether malicious acts are suspected and whether cross-border impact is possible. You can send it with partial facts. Article 23 explicitly sequences the detail across the later reports. For the deadlines in depth, see the 24-hour, 72-hour and one-month deadlines.
A classification routine that works in 30 minutes
Build the decision into the incident plan before the incident. Article 21(2)(b) incident handling already requires a process. Add a one-page classification step:
- Which service is affected, and what is the realistic worst case in the next 24 hours, not the current state?
- Could the financial loss reach the sector or national threshold? Use the lower of the fixed figure and the turnover percentage where CIR 2024/2690 applies.
- Is any other person or entity, including a client of the client, exposed to considerable damage?
- Does a sector-specific criterion apply?
- Has the same root cause produced another incident in the last six months?
Write down the answer and the time. A dated classification record, even for a "not significant" call, is the evidence a supervisor asks for first.
If the answer is unclear, note that Article 30 lets entities report incidents voluntarily. The mere act of notification does not subject the entity to increased liability, as Article 23(1) states. Our guide to voluntary near-miss notification covers when that is the sensible route.
Why the classification record matters later
Supervisors do not only ask whether you reported. They ask how you decided. Under Article 32, essential entities can face on-site inspections and security audits, and a classification log is the cheapest document that shows the incident process actually ran. It also protects you commercially: if a client later disputes whether an incident should have been reported, your dated reasoning shows the call was made on stated criteria, not in a panic. Keep the log with the incident ticket, and review it quarterly against the incidents that were not reported.
Test your clients before the next Friday night
Most clients have an incident plan that says "report significant incidents" and no definition of the term. Ask each one three things: who decides, against which thresholds, and where it is written down. A NIS2Certify quick scan shows in minutes whether the incident handling measure has evidence behind it, and gives you the gap list to work from.
The practical takeaway: define "significant" per client, in writing, with numbers, before the 16:40 incident arrives.
Still have a question?
Answers are generated from our articles and are not legal advice. Do not enter personal or confidential data.
