ISO 27001 Control A.5.28: Collection of Evidence
When something goes wrong, the first instinct is to fix it. Wipe the laptop, reset the account, restore the server. Those instincts are right for restoring the business, and wrong for understanding what happened. A.5.28 asks you to think about evidence before you need it, so that the facts survive the clean-up. It is the 28th of 37 organizational controls in our walk through Annex A.
You might never go to court, and most incidents never come close. But evidence matters well beyond litigation: it supports insurance claims, regulator inquiries, disputes with suppliers, disciplinary action and, most simply, an honest root cause analysis. Evidence that was lost in the first hour cannot be recovered in the second week.
What the control requires
Annex A control 5.28 of ISO 27001:2022 states that the organization shall establish and implement procedures for the identification, collection, acquisition and preservation of evidence related to information security events.
The key elements:
- Procedures - written, agreed steps, not something improvised by whoever is on call
- Identification - knowing what could be relevant: logs, devices, emails, access records, images
- Collection and acquisition - gathering it in a way that does not change it
- Preservation - keeping it safe, intact and traceable afterwards
- Related to information security events - covering events you have not yet classified, as well as confirmed incidents
A.5.28 is a detective control in the attributes of ISO/IEC 27002:2022. Annex A states the control; ISO 27002 gives guidance, including documented internal procedures for handling evidence for disciplinary and legal purposes, a chain of custody, protection of evidence, and the involvement of competent people. It also notes that requirements differ by jurisdiction, so we suggest taking legal advice about what standard your evidence would have to meet where you operate.
How to implement evidence collection
Decide what you may need and where it lives
Make a short list of the sources that usually matter in an incident, and check that they actually exist and are retained long enough to be useful. Typical sources include:
- System, application and authentication logs, including cloud service audit logs
- Email headers and message copies relating to phishing or data leaks
- Endpoint images or memory captures from affected devices
- Network records, firewall and proxy logs
- Physical records such as door access logs and camera footage from the physical security monitoring systems
- Records of who accessed what, from your access control systems
For each source, note how long it is kept, who can export it and how. A common surprise is that a cloud service keeps audit logs for 30 days by default and nobody has extended that. Learn this before the incident, not during it.
Write a simple collection procedure
Your procedure should be short enough to follow under stress. Cover the following points:
- Do no harm. Prefer copying data to altering it. Isolate rather than power off where practical, because volatile data such as memory is lost when a machine shuts down.
- Record everything. Note who collected the item, when, from where, by what method, and how it was stored. Use a form or a table with a line for each item.
- Make verified copies. Where you copy digital evidence, record a hash value so that you can later show the copy has not changed.
- Work on copies. Analysis should use the copy, with the original stored securely.
- Limit access. Only people who need to handle the evidence should have access, and each handover should be logged.
Keep the chain of custody log together with the incident record produced under response to information security incidents. If the response is already running, a single timeline can hold both response actions and evidence handling.
Know your limits and bring in competent help
Proper forensic collection takes skill. Some tasks, such as imaging a disk or capturing memory from a live server, are best done by someone who has done them before. Decide in advance who that would be: an internal specialist, a trusted provider or your incident response partner. For ordinary cases, such as exporting logs or preserving an email thread, trained staff can follow the procedure. For anything that might become a legal dispute, or involve law enforcement or a regulator, hand over to competent people early and preserve everything as it is. Contact with authorities explains how to establish those channels before you need them.
Be mindful of privacy and employment rules. Collecting evidence from an employee’s device, mailbox or private communications raises questions that depend on local law, your policies and the notices you have given staff. We suggest agreeing the approach with legal or HR advice, and making sure your terms and conditions of employment and acceptable use rules set the expectations clearly.
Secure the evidence and the process
Store evidence in a protected location with restricted access, ideally with logging of who opens it. Use encryption where the content is sensitive. Define how long evidence is retained, and what happens to it afterwards, balancing the need to keep it for a possible dispute against data protection duties to avoid keeping personal data longer than necessary. Include suppliers in the thinking: if an incident involves a cloud provider, find out what they can supply and how fast, as described in information security for use of cloud services.
Like every other part of the incident process, evidence handling improves with rehearsal. During a tabletop exercise, add a step where the team has to preserve evidence from a compromised laptop or export audit logs from a cloud tenant, and see what is missing. Typical discoveries include forgotten administrator rights, short log retention and nobody knowing where the hashing tool is installed. Fix those, then repeat the exercise a few months later. The goal is not courtroom-grade perfection for every event, but a team that instinctively pauses before wiping a machine and asks whether anything needs capturing first.
Related controls and clauses
A.5.26 connection. Response to information security incidents runs alongside evidence collection; containment should be chosen with evidence in mind.
A.5.27 connection. Learning from information security incidents depends on having accurate facts to learn from.
A.5.25 connection. Assessment and decision on information security events happens early, when evidence is most at risk of being lost.
A.6.4 connection. The disciplinary process relies on credible, properly handled evidence when conduct is at issue.
Clause 7.5.3 connection. Control of documented information principles of protection, retention and access apply to evidence records too.
Forward reference. The technological controls on logging (A.8.15) and clock synchronization (A.8.17) supply much of the digital evidence this control collects, so we return to them later in the series.
What auditors check
A documented procedure. Auditors look for written steps covering identification, collection and preservation.
Chain of custody records. For past incidents, they check whether items were logged, who handled them and how they were stored.
Availability of sources. They ask how long key logs are kept and whether the retention matches what the procedure assumes.
Competence. They ask who would collect evidence and whether those people have the skills or an arrangement with someone who does.
Integrity measures. Hashing, restricted access and protected storage show that evidence would hold up to scrutiny.
Alignment with legal advice. Where legal or regulatory requirements apply, they look for evidence that these were considered.
Common mistakes to avoid
Wiping first. Rebuilding an affected machine before capturing anything is the classic error.
Short log retention. If logs vanish after a few days, the evidence is gone before anyone realizes there is an incident.
No record of handling. Evidence without a trail can be challenged, even if nobody tampered with it.
Doing forensics by amateur. Poorly handled collection can destroy the very data you hoped to preserve. Know when to call for help.
Ignoring employment and privacy rules. Collecting from personal accounts or devices without a clear basis can create new problems. Get advice first.
Keeping evidence forever. Indefinite retention of personal data in evidence stores is a risk of its own.
How 27kay can help
We help organizations write evidence procedures that are proportionate and practical, and make sure the underlying logs and retention are in place. As part of ISO 27001 implementation, we help you define the steps, test them in an exercise and prepare the records an auditor will ask for. For the full picture, see our ISO 27001 knowledge hub.
If you needed proof of what happened on a server last Tuesday, would the logs still exist? Get in touch - we will help you find out while it is still a question and not a crisis.