Incident response plan: steps, roles and the EU reporting clocks
An incident response plan is the written answer to one question: when something goes wrong with your systems or data, who does what, in which order, and who must be told by when. In the EU, that last part now has hard deadlines. NIS2 gives in-scope companies 24 hours for an early warning, the GDPR gives 72 hours for a personal data breach, and the Cyber Resilience Act gives product makers 24 hours for an actively exploited vulnerability. This guide shows how to build a plan that meets those clocks, with sources read on 12 November 2026.
Some companies call the same document an IT emergency plan, a cyber incident response plan or a security incident procedure. The name matters less than three qualities: it is short, it names real people, and it has been tested.
Why a written plan
Most of the damage in an incident comes from the first hours: systems left connected while ransomware spreads, evidence wiped by a well-meant reinstall, a reporting deadline missed because nobody knew it existed. A written plan replaces improvisation with decisions taken calmly in advance. It is also required. NIS2 lists incident handling as a mandatory measure (Article 21(2)(b)), together with business continuity and crisis management (Article 21(2)(c)), and customers increasingly ask suppliers to show their plan in tenders.
The reporting clocks
| Law | Who | What and when |
|---|---|---|
| NIS2, Article 23(4) | Essential and important entities | Significant incident: early warning within 24 hours, incident notification within 72 hours, final report within one month of the notification |
| GDPR, Article 33 | Any controller of personal data | Personal data breach: notify the supervisory authority without undue delay and, where feasible, within 72 hours, unless unlikely to result in a risk to people |
| GDPR, Article 34 | Any controller | High risk to people: inform the people affected without undue delay |
| Cyber Resilience Act, Article 14 | Manufacturers of products with digital elements | Actively exploited vulnerability: early warning within 24 hours, notification within 72 hours, final report within 14 days of a fix; applies since 11 September 2026 |
Each clock starts when you become aware, not when the investigation is finished. One incident can start several clocks at once: a ransomware attack at an in-scope company that also exposes customer data triggers both NIS2 and GDPR reports, to different authorities. The plan should name who files each report and keep ready-made templates. Under GDPR Article 33(2), a processor such as a software or hosting provider must notify its controller without undue delay, so your contracts with suppliers belong in the plan too. More on NIS2 is in our NIS2 compliance guide, and on product duties in our Cyber Resilience Act guide.
The steps of incident response
Frameworks name the phases differently, but the work is the same. NIST SP 800-61 Revision 3, published in April 2025 and replacing Revision 2 from 2012, organises incident response around the NIST Cybersecurity Framework 2.0, from preparation through detection, response and recovery. In practice, a plan covers six steps:
- Prepare. Roles, contacts, asset list, logging, backups, access to admin accounts in an emergency, report templates, and a list of external help such as forensics, legal counsel and the insurer.
- Detect and assess. How an incident is noticed (alerts, staff reports, a customer, a supplier), who confirms it, and how its severity is rated. The severity scale decides who is called and whether a reporting clock starts.
- Contain. Stop the spread without destroying evidence: isolate affected devices from the network rather than switching them off, disable compromised accounts, block malicious addresses, and preserve logs.
- Eradicate. Remove the cause: malware, the backdoor account, the unpatched vulnerability, the stolen credentials. Rebuild from known good sources where trust is lost.
- Recover. Restore systems and data from backups that are checked to be clean, bring services back in priority order, and watch closely for signs of return.
- Learn. A short review within a few weeks: what happened, what worked, what failed, and a list of improvements with owners and dates. NIS2's final report asks for the root cause and mitigation, so this step also feeds the authorities.
Roles and contacts
An incident is the wrong moment to find out who decides. The plan names, with deputies:
- Incident lead: runs the response, decides on containment and escalation, keeps the log.
- Technical responders: one per critical system, including external suppliers who run them.
- Communication: staff, customers, partners and, if needed, the press, with pre-approved wording.
- Legal and data protection: decides whether a reporting duty applies and files the reports.
- Management: approves spending, external help and decisions such as shutting down a service. Under NIS2 Article 20, management also approves the measures and can be held liable.
Keep the contact list outside the systems it depends on. If email and the file server are encrypted by ransomware, a plan stored only there is of no use. A printed copy and a copy in a separate account solve that.
An incident response plan template outline
- Purpose, scope and the date of the last review.
- Definitions: what counts as an incident, and the severity levels with examples.
- Roles, names, deputies and contact details, including suppliers and external help.
- The steps, from detection to the post-incident review, with a decision point for each reporting duty.
- Reporting: which authority, which deadline, which template, who files.
- Communication: internal and external, with pre-approved messages.
- Evidence: what to preserve and how, and a running incident log.
- Playbooks for the most likely incidents, one page each.
- Links to the business continuity plan, the backup and restore procedures, and the asset list.
- Testing schedule and the record of past exercises and incidents.
Test it before you need it
A plan that has never been used will fail in small, avoidable ways: the backup restore takes three days instead of three hours, the admin password is with someone on holiday, the authority's reporting portal needs a registration nobody made. A yearly tabletop exercise, where the team walks through a realistic scenario around a table, finds these gaps cheaply. A restore test from backup, with the time it took written down, should happen at least as often. After each exercise or real incident, update the plan and record what changed.
What software can do for incident response
Good incident response depends on the systems being ready for it. Central logging with enough history to reconstruct what happened, alerts on the events that matter, MFA on every admin account, an asset list that is current, backups that are separated from the production network, and a way to disable a user everywhere in one step. In your own software, an audit log of who did what and when, and a feature switch to turn off a compromised function quickly, are worth building in from the start. A simple incident register, where each incident gets a timeline, decisions, deadlines and reports, keeps the reporting clocks visible.
How we help
We write and test incident response plans with your team, set up the logging, backups and access controls that make them work, and build incident registers and audit logs into custom systems. A security audit shows the gaps first. For ongoing operation, managed IT services include monitoring, patching and tested restores, and co-managed IT supports an internal team. If your company needs a confidential channel for staff to raise concerns, see our whistleblowing software.
Questions people ask
What is the difference between an incident response plan and a business continuity plan?
The incident response plan covers a security incident: how it is detected, contained, investigated, reported and closed. The business continuity plan covers how the business keeps working while systems are down, whatever the cause, from a cyberattack to a power cut or a supplier failure. The two meet in recovery, so the incident response plan should point to the continuity plan rather than repeat it.
How long should an incident response plan be?
Short enough to use under stress. A main document of a few pages with roles, contacts, severity levels and the steps, plus one-page playbooks for the incidents you expect most, such as ransomware, a compromised account, a lost laptop or a data leak. Long background material belongs in an appendix.
Who should be on the incident response team?
A lead who decides, someone technical for each critical system, a person for communication, someone who knows the legal reporting duties and data protection, and a member of management who can approve spending and external help. In a small company one person may hold several roles, but every role needs a named deputy.
How often should the plan be tested?
At least once a year, and after any major change to systems, suppliers or staff. A tabletop exercise of two hours, walking through a realistic scenario, finds most gaps: missing phone numbers, unclear decisions, backups nobody has restored. Record the date and the findings.
Is there a standard template we can follow?
NIST Special Publication 800-61 Revision 3, published in April 2025, gives incident response recommendations organised around the NIST Cybersecurity Framework 2.0 and is free to read. A template is a start; the plan only works once the names, systems and reporting routes are your own.
Sources
- Directive (EU) 2022/2555 (NIS2), Articles 20, 21 and 23 (read 12 November 2026)
- Regulation (EU) 2016/679 (GDPR), Articles 33 and 34 (read 12 November 2026)
- Regulation (EU) 2024/2847 (Cyber Resilience Act), Articles 14 and 71 (read 12 November 2026)
- NIST SP 800-61 Rev. 3, Incident Response Recommendations and Considerations for Cybersecurity Risk Management (read 12 November 2026)