AnySec
Incident Response Report: What It Must Contain — AnySec Insights
← InsightsResponse · 8 min read

Incident Response Report: What It Must Contain

A narrative of what happened is not an incident response report. The sections to require, the evidence a regulator or insurer can re-check, and the red flags.

By AnySec engineering

The short answer

A useful incident response report gives a timeline in one time zone, the scope of what was affected, a root cause stated with its evidence or an honest unknown, every action taken and who authorised it, how evidence was preserved, what remains open with owners, and what the investigation could not cover. A narrative of what happened, with no way to check it, is a summary rather than a deliverable.

Who this is for

This is for a CTO, head of security or risk owner who holds, or is considering, an incident response retainer or a one-off engagement and wants to know what the document at the end must prove. It is equally for whoever must hand that document to a regulator, an insurer, a payment partner or a board. Before the report there is the contract: what an incident response retainer quote should say covers hours, overage and the response clock. Before the incident there is rehearsal, which the tabletop exercise guide covers. This article is about the artefact that outlives both: the report.

Why a narrative is not a report

Disappointing incident reports fail in a few predictable ways.

  1. They tell a story with no evidence trail. Each claim, such as "the attacker entered through the VPN on Tuesday", should point to a log, an image or a record you can check. Otherwise the reader must trust the author.
  2. They blur what was known with what was assumed. An investigation under time pressure rests on hypotheses. A report should label which were confirmed, which were ruled out and which were left open.
  3. They omit decisions. Containment often means a business trade-off: take a service offline, rotate a key, suspend withdrawals. The report should record who decided, on what information and when.
  4. They are written for one audience. The engineer needs hosts and indicators. The board needs impact and decisions. A regulator needs scope and timing. One document rarely serves all three.
  5. They end at containment. Recovery steps, remaining exposure and the changes that would shorten the next response are the part that actually reduces future loss.

Because the report is read long after the people involved have moved on, its structure belongs in the quote, not in a discussion after the fire is out.

A hand places a sealed evidence pouch with a single lime-green seal into a custody tray beside a printed timeline sheet on a dark desk
Every claim in the report should trace to evidence that was sealed, logged and handed over, not remembered.

The report anatomy matrix

Use this as a checklist when you write the quote or review a draft. It contains no thresholds or timings, because the right ones come from your contract, your licences and your own risk.

SectionWhat it must showRed flag if missingWho relies on it
Executive summaryWhat happened, impact and status in plain language, with the confidence levelTechnical jargon only, or reassurance with no basisBoard, executives
TimelineEvents from first sign to closure, in one time zone, each with a sourceDates without sources, mixed time zonesEveryone
Scope and impactSystems, accounts, data classes and customers affected, and how that was established"Limited impact" with no methodLegal, regulator liaison
Detection and entryHow the incident was found and the most likely way in, labelled confirmed or hypothesisA single confident cause with no evidenceSecurity lead
Actions and decisionsEach containment and recovery action, who authorised it, and the timePassive voice, no names or rolesRisk owner, auditor
Evidence registerWhat was collected, from where, by whom, how it was hashed and stored"Evidence was preserved" with no listLegal, insurer
Containment statusWhat is contained, what is not, and what is being monitoredImplied full closureOperations
Remediation planFixes with owner, priority and verification methodRecommendations with no ownersEngineering
Limits and open questionsWhat the investigation could not determine and whySilence, which looks like completenessEveryone
Notification inputsThe facts needed for any required notifications, kept separate from legal conclusionsThe provider deciding what you must discloseLegal, compliance

The two rows that separate a real report from a story are the evidence register and the limits section. One shows the work can be re-checked; the other shows the author is willing to say where it stops.

The timeline: the part everyone reads first

A timeline is easy to write badly. Ask for these properties.

  • One time zone, stated at the top, so log entries, vendor tickets and customer reports line up.
  • A source on every line: the log, image, ticket or interview behind the entry.
  • Separate columns for attacker activity, detection and response, so you can see how long each stage lasted without inferring it.
  • Gaps shown as gaps. If logs were missing for a period, the timeline should show the hole rather than interpolate across it.

For regulated operators, the same timeline is often the basis of a regulatory or card-scheme notice, so its accuracy matters beyond the engineering team. The digital bank payment-fraud response clock shows how those external clocks sit alongside the technical one.

Actions and decisions: who authorised what

Incident response involves actions with real cost: taking a customer-facing system offline, rotating credentials, blocking a payment route, suspending a market. The report should record each material action with:

  1. What was done, in a form someone could reverse or repeat.
  2. Who authorised it, by role, and who carried it out.
  3. What was known at the time, so the decision can be judged fairly later.
  4. What it cost or risked, such as downtime or lost transactions, without invented figures.

This section protects both sides. It lets your auditors see that decisions were deliberate, and it stops a later review from judging a decision with information that arrived hours afterwards.

Evidence an insurer or regulator can re-check

A report is only as strong as what stands behind it. Ask for these to be named, even if the raw material is held separately.

  1. An evidence register listing each item, source system, collection time, collector and storage location.
  2. Integrity records, such as hashes recorded at collection, so an image can be shown unchanged.
  3. Chain-of-custody notes for anything that changed hands, including to a third party or a law-enforcement contact.
  4. Retention terms: who keeps the evidence, for how long and under what access rules, because the questions rarely arrive during the incident.
  5. The method behind scope and impact claims, such as which logs were searched and for what, so a second examiner could repeat it.

The sector-specific evidence differs: a sportsbook must preserve bet and market state, and a lottery must preserve draw records. Your report should reflect the evidence your own regulator will ask for.

Three audiences, three versions

Decide before an incident who needs what. A practical split:

  • Technical report. Hosts, indicators, commands, configuration, full timeline. Audience: your security and engineering team, and any later examiner.
  • Management summary. Impact, decisions taken, current status, what remains open and what it will cost to fix. Audience: executives and the board.
  • Redacted external version. Facts needed by a regulator, insurer, partner or card scheme, with sensitive indicators removed. Audience: outside parties.

Ask whether the provider writes all three from one evidence base, who approves each, and whether they will be available to answer questions from an auditor or insurer afterwards. If a card-data breach is possible, check whether a PCI forensic investigator is needed instead, since that role has its own reporting format.

Red flags when you review a draft

  • A confident root cause with no evidence references.
  • A timeline with mixed time zones or no sources.
  • "Limited impact" without a stated method for establishing scope.
  • Actions listed without who authorised them.
  • No evidence register, or no hashes or custody records.
  • No limits section, or limits buried in an appendix.
  • Recommendations with no owner or verification step.
  • The provider drafting your legal conclusions or notifications rather than supplying facts to your counsel.
  • A rewrite of conclusions after the client objected, rather than a factual correction.

Ask for the report before you sign

Treat the deliverable as a quote line. Before an incident, ask for a redacted sample from a past engagement, the headings the provider will use, the delivery points (a verbal brief, an interim summary, the final report), who may edit the draft, who keeps the evidence and for how long, and whether report time sits inside the retainer hours or is billed separately. If no sample is available, treat that as information. A report also does not replace detection: ongoing coverage is the job of managed SOC, and a recurring root cause is a signal for a penetration test or a security hardening review.

Limitations

This article is not legal or regulatory advice and does not state what any regulator, insurer, card scheme or licensing authority requires in an incident report or by when. It uses no external statistics and no incident case studies. The matrix is a review checklist to adapt to your own contract, jurisdiction and estate, and it contains no thresholds or timings. An incident report describes what was found within the limits of the evidence available at the time; it does not prove that an attacker is gone, that no other system was touched, or that a similar incident will not recur. Reports can change as new evidence arrives, so the version and date matter.

Decision and next step

If you can read the draft and, for each claim, find its source, for each action find who authorised it, and for the limits find a clear statement of what is unknown, the report is doing its job. If not, ask for the evidence behind the summary or reject the deliverable. To have the report, its three versions and its evidence rules written into the engagement before an incident, get an IR retainer from AnySec and we will map your estate, your notification obligations and the evidence you would need to the sections you receive.

Sources and review

The framing follows public guidance: NIST SP 800-61 Rev. 3, Incident Response Recommendations and Considerations for Cybersecurity Risk Management, which covers incident response lifecycle activities and lessons learned, and NIST SP 800-86, Guide to Integrating Forensic Techniques into Incident Response, which covers evidence collection and handling, read 2026-11-19; check the current versions, and read your own regulator's and insurer's reporting requirements. AnySec prepared the report matrix from the public structure of incident response engagements; no AnySec-original statistics are used. Author: AnySec Engineering. Published 2026-11-19; reviewed by the AnySec security team.


Related reading

To have the deliverable specified up front, get an IR retainer.

Frequently asked questions

What should an incident response report contain?
A timeline in one time zone, the scope of affected systems and data, the confirmed root cause or an honest statement that it is unconfirmed, the actions taken with who authorised them, the evidence held and how it was preserved, what was contained and what was not, remediation with owners, and the limits of the investigation. A story of the incident without evidence references is a summary, not a report.
Who should receive the incident response report?
Decide before the incident. The technical report goes to your security and engineering leads. A shorter management version goes to executives and the board. A third, redacted version may be needed for regulators, insurers, partners or card schemes. Ask the provider whether they produce all three from one body of evidence, and who controls distribution.
How is an incident response report different from a forensic report?
An incident response report covers the whole engagement: detection, containment, eradication and recovery decisions. A forensic report goes deep on evidence, methods and findings from specific systems and is written to be examined by third parties. Many engagements need both, and a quote should say which is included.
What if the root cause is never confirmed?
Then the report should say so, list the hypotheses tested, state what evidence was missing, such as logs that were not retained, and name what you should change so the next investigation can answer. A confident root cause with no supporting evidence is worse than an honest unknown.
When do we receive the report?
Ask for the delivery points in the quote: a verbal brief at the end of containment, a written interim summary, and the final report after evidence review. Also ask who may edit the draft. A provider should accept factual corrections but not rewrite conclusions.

Rather not learn this in production.

Talk to the engineers behind these write-ups — no sales script, a straight read on where you stand.

Get a fixed quote