ISO 27001 Control A.5.26: Response to Information Security Incidents

· 8 min read · Lyudmil Arkov

This is the moment the earlier controls were preparing for. An event has been assessed, someone has said “this is an incident”, and now people have to act. A.5.26 asks that they act according to documented procedures rather than improvising. It is the 26th of 37 organizational controls in our walk through Annex A, and the middle step in the incident life cycle.

Improvisation feels natural under pressure, and it is how well-meaning people make incidents worse: wiping a machine that held the evidence, emailing a customer before the facts are known, or restoring from a backup that already contained the intruder. A short, tested procedure keeps the response calm and gives everyone the same picture.

What the control requires

Annex A control 5.26 of ISO 27001:2022 states that information security incidents shall be responded to in accordance with the documented procedures.

The key elements:

  • Information security incidents - events that your triage has classified as incidents
  • Responded to - containment, investigation, communication, eradication and recovery, not only noticing
  • Documented procedures - written steps, available to the people who need them
  • In accordance with - the procedures are followed in practice, and deviations are explained

A.5.26 is a corrective control in the attributes of ISO/IEC 27002:2022. Annex A states the control; ISO 27002 gives guidance, including containing the incident, collecting evidence, escalating where needed, making sure all response activities are logged, communicating on a need-to-know basis and closing the incident formally once it is resolved. The procedures should be the ones you defined under information security incident management planning and preparation.

How to implement incident response

Contain first, then investigate

The first priority is to stop the damage from growing. Depending on the incident, that might mean disabling an account, isolating a laptop from the network, revoking a leaked key, blocking an address or taking a service offline. Write containment options into the procedure for the incidents you consider most likely, with a clear note on who may authorize each one. Prefer actions that preserve evidence: isolating a machine is usually better than switching it off or rebuilding it, because memory and logs may still hold the answers.

Weigh the business impact before acting. Taking down a revenue-generating service is a serious step, and the incident lead should be able to explain the choice afterwards. That is why decision rights matter, as we describe in information security roles and responsibilities.

Keep a running log

From the first minute, someone should record what is known, what has been done, by whom and when. A simple timeline in a shared document is enough. This log serves several purposes: it keeps the team coordinated when people hand over, it supports later review, and it provides the factual basis for any notification to customers or authorities. Write observations and actions separately from guesses, and mark which is which. If the incident might involve legal action or a regulator, the log is part of your record of events, and the guidance in control A.5.28, collection of evidence, will apply. We cover that control in detail shortly, but the habit starts now: record times accurately, note who handled what, and avoid altering original data.

Communicate with care

Decide who tells whom, and when. Internally, keep the circle informed but small, and use a channel that an attacker cannot read if your normal systems might be compromised. Externally, customers, partners and authorities may need to hear from you. Duties and deadlines depend on your sector, contracts and jurisdictions, so we recommend agreeing them with qualified advisers in advance and writing them into the procedure, including who approves the wording. See contact with authorities for the relationships worth establishing before you need them.

Be factual and avoid speculation in messages. If you do not yet know the cause, say so and promise an update. Customers usually react better to a calm, honest note than to silence followed by a polished explanation.

Recover, verify and close

Eradication means removing the cause, not only the symptoms: closing the vulnerable entry point, resetting credentials, rebuilding affected systems from known good sources. Recovery means restoring services and checking that they behave as expected. Before you declare the incident closed, check that the intruder or the fault is really gone, that monitoring is watching for a return, and that the affected data owners have been told. Then close the record with a summary of what happened, the impact, the actions taken and any follow-up work. Open items go into your corrective action list, which is the subject of nonconformity and corrective action.

Not every organization has the skills to investigate a serious intrusion. Decide in advance when you would call for outside help, such as signs of ransomware or access by an unknown party to sensitive data, and who you would call. A retainer or at least a prepared contact list means you are not searching for a provider while the incident is unfolding. When an outside party is involved, agree on how evidence and access are shared and what confidentiality terms apply.

A lost laptop, a phishing compromise, a ransomware infection and a misdirected email are all incidents, but they need different first moves. Rather than one generic procedure, keep a short playbook for each of your most likely scenarios. A good playbook fits on one page and answers four questions: how do we recognize this, what do we do in the first hour, who must be told, and what needs checking before we close it. For a lost device, the first hour might be remote lock or wipe, credential reset and a check of what data was on it. For a phishing compromise, it might be password reset, session revocation, mailbox rule review and a search for other recipients. Start with three or four playbooks based on what your information security risk assessment says is most likely, and add more as real incidents teach you what you are missing. Keep them where the team can reach them without your main systems.

A.5.24 connection. Information security incident management planning and preparation provides the roles, contacts and process that the response follows.

A.5.25 connection. Assessment and decision on information security events determines which events enter the response and at what priority.

A.5.5 connection. Contact with authorities covers the external relationships used during serious incidents.

A.5.22 connection. When a supplier is involved, monitoring, review and change management of supplier services and your supplier agreements determine how they must support you.

A.6.8 connection. Information security event reporting is how many incidents are first spotted.

Clause 10.2 connection. Nonconformity and corrective action takes the follow-up actions from incidents into the management system.

Forward reference. Learning from information security incidents (A.5.27) turns each response into improvement, collection of evidence (A.5.28) covers preserving proof, and information security during disruption (A.5.29) addresses larger interruptions. We cover these in the next posts.

What auditors check

Documented procedures. Auditors expect written response procedures that match the incident types you face.

Records of real incidents. They sample actual incident files and compare what happened with what the procedure says should happen.

Timeline and ownership. They look for logs showing who did what and when, and for a named owner for each incident.

Communication. They check how stakeholders were informed and whether any notification duties were considered.

Closure. They look for formal closure with a summary and follow-up actions rather than incidents that simply fade away.

Exercises. If there have been few real incidents, they ask for evidence of tests or simulations, and whether the findings were acted on.

Common mistakes to avoid

Rebuilding too early. Wiping a system before collecting what it can tell you removes the evidence needed to understand the cause.

One hero. A response that relies on a single person’s knowledge collapses when that person is away.

No log. Without a timeline, you will struggle to explain the incident to management, customers or an auditor.

Over-sharing. Posting details in a broad chat channel can leak sensitive facts or tip off an attacker who is reading your systems.

Declaring victory too soon. Closing an incident without checking for persistence invites a repeat.

Forgetting suppliers. If the incident originates with or affects a provider, make sure your contract gives you the support you need.

How 27kay can help

We help organizations turn incident response from a document into a habit. As part of ISO 27001 implementation, we review your procedures, help you rehearse realistic scenarios and prepare the records an auditor will want to see. For the full picture, see our ISO 27001 knowledge hub.

If a serious incident started this afternoon, would your team follow a plan or make one up as they go? Get in touch - we will help you build the kind of response you will be glad to have used.