What incident response is

Events, incidents, and evidence, and why forensics belongs in the response

By the end of this lesson you can

  • Tell an event, an adverse event, and an incident apart, using the NIST definitions in your own words.
  • Give three examples of incidents that could hit a small organization.
  • Explain why evidence matters even when nobody ever goes to court.
  • Describe how incident response and digital forensics fit together, and name the goals of a response.

01A Tuesday at Northwind Dental

Throughout this course we follow a fictional example: Northwind Dental, a small dental practice with eight staff, two treatment rooms, a reception desk, and a patient booking portal at portal.example.com. Nora is the practice manager. Sam is the part-time IT generalist. Alice works at reception on the PC called RECEP-01. You will spend most of the course in Sam's shoes.

On a Tuesday morning at 08:40 local time, Alice signs in to the booking portal and opens the page that lists recent logins to her account. One entry says her account signed in at 06:12 from the address 198.51.100.23. Alice was asleep at 06:12, and nobody at the practice uses that address. She calls Sam.

Sam now faces the questions this course is about:

  • Is this an incident, or just something odd with an innocent explanation?
  • Who needs to know, and how quickly?
  • What must be written down or preserved before anyone starts changing passwords or clicking around?
  • How will Northwind later know what really happened, and whether patient data was seen?

A calm, practiced answer to these questions is the difference between a bad morning and a bad month. Panic tends to destroy exactly the traces you need, while a simple routine protects them.

02Events, adverse events, and incidents

NIST, the US National Institute of Standards and Technology, publishes a widely used guide on the topic: SP 800-61 Revision 3, released in April 2025. It uses three words that build on each other. Here they are in plain language.

  • An event is anything observable that happens on computers, networks, services, or cloud systems: a login, a file save, a software update, a printer going offline. A busy office produces many thousands every day, and almost all of them are harmless.
  • An adverse event is an event with a negative consequence, whatever the cause: a crash, a failed backup, a burst of failed logins. Some are accidents, some are attacks. Nobody knows which until someone looks.
  • A cybersecurity incident is an occurrence that actually or imminently puts the confidentiality, integrity, or availability of information or systems at risk without lawful authority, or that breaks (or is about to break) the law, security policies, procedures, or acceptable use rules.

Confidentiality means only the right people can read the data. Integrity means the data has not been changed without permission. Availability means people can use the system when they need it. An incident threatens at least one of the three, or breaks a rule.

A funnel: many events at the top, fewer adverse events in the middle, a few incidents at the bottom, with a side arrow saying analysis decides which events move down Events many thousands a day Adverse events negative effect Incidents a few Analysis decides
Most events are harmless. Only analysis tells you whether an adverse event is part of an incident.

NIST's examples of incidents, in our words, show how varied they are:

  • An attacker makes a web service unavailable so that customers cannot reach it.
  • Someone steals administrator credentials at a software-as-a-service provider and uses them to read customer data.
  • Ransomware encrypts an organization's files after the attackers have copied data out.
  • A phishing email tricks an employee, and the attacker uses the access to commit fraud.

Now try it on Northwind. The table below classifies six situations. Before you open it, decide for yourself, then compare.

Show the classification of the six situations
Situation (fictional)Event?Adverse event?Incident?
Alice mistypes her password once, then logs inYesNo, nothing bad happenedNo
A power outage stops all PCs for an hourYesYes, systems unavailableUsually no: handled by continuity plans, unless tampering is found
A successful login to the portal from an unknown countryYesLikely, if the user did not make itSuspected: analysis decides, treat it seriously
A USB stick found in the parking lotYesPossiblyNot by itself, but plugging it in could start one
A mass email with a link arrives in every inboxYesYes, if the link is maliciousPossibly: it depends on whether anyone clicked or typed a password
Files on the shared drive are suddenly encryptedYesYesYes, availability (and maybe confidentiality) is hit
Myth

"An incident is only a hack by outsiders." The definition does not mention outsiders at all. A staff member who copies patient lists to a personal account breaks policy, and that is an incident. So is an administrator who accidentally makes a folder of records public. Intent and origin matter for the response, not for the label.

03What incident response tries to achieve

Incident response is the organized way an organization detects, analyzes, contains, removes, and recovers from incidents, and learns from them. "Organized" is the key word: decisions, roles, and records are planned in advance, so that nobody has to invent a process at 06:30 on a Tuesday.

A response has four goals, which NIST's guide covers across its whole structure:

  1. Limit the damage. Stop the attacker from going further, and stop data from leaking further. At Northwind that could mean locking Alice's portal account and checking whether other accounts were used.
  2. Restore normal operations. Patients must be able to book appointments again, with systems you can trust.
  3. Find the root cause. How did the attacker get Alice's password? A reused password, a phishing email, malware on RECEP-01? If you fix only the symptom, the same door stays open.
  4. Improve. Every incident, and every near miss, teaches something: a missing log, a slow phone call, a weak password rule. Writing it down and fixing it is part of the response, not an optional extra.

These goals can pull against each other. Restoring quickly by wiping a PC destroys the traces that would reveal the root cause. Watching an attacker to learn more lets the damage continue. Later lessons show how to weigh these choices and who should make them.

04Where forensics comes in

Digital forensics is the careful collection, examination, and analysis of digital data so that the findings can be trusted, explained, and repeated by someone else. In incident response it answers the hard questions: what happened, when, through which door, and what was touched.

NIST SP 800-86, the guide on integrating forensic techniques into incident response, describes four phases. They give this course its shape:

  1. Collection: identify possible sources of data, label and record them, and acquire the data while protecting its integrity.
  2. Examination: process large amounts of collected data and extract the parts that matter.
  3. Analysis: draw conclusions that answer the questions of the case, with methods you can justify.
  4. Reporting: explain what was done, which tools were used, what was found, and what should happen next.

Everything you collect is evidence: data that may support or disprove a statement about what happened. NIST's Rev. 3 makes a practical point here. Not every incident needs a formal chain of custody, the signed record of who held an item when, but the data you collect is still evidence, and its integrity and origin should be preserved.

Why evidence matters even without a court

Most incidents in small organizations never reach a courtroom. Evidence still matters, for very ordinary reasons:

  • Decisions depend on it. Whether Northwind must tell patients or an authority depends on whether patient data was actually reached. "We think not" is weak, and "the logs show only the calendar page was opened" is strong.
  • Others will ask. An insurer, a provider, or a data protection authority may ask what happened and how you know.
  • You may be wrong at first. Early theories often change. Good records let you test a new theory without starting again.
  • Things can escalate. A case that starts small can later involve police or lawyers. Evidence handled carelessly on day one cannot be repaired on day ten.
Defender's view

Before touching anything, write down the first observation with the date, the time, and the time zone, who saw it and exactly what they saw. Sam's first line could read: 2026-03-10 08:40 CET (07:40 UTC), Alice, RECEP-01: portal shows login to her account at 06:12 CET from 198.51.100.23, not made by her. Systems often log in UTC while people think in local time, so a missing time zone can shift the whole story by an hour or more.

05Who responds in a small organization

NIST lists many roles that take part in a response: leadership, incident handlers, technology staff, legal advisers, public affairs and media relations, human resources, physical security and facilities, and the owners of the affected assets. Third parties such as cloud and internet providers also play a part, with duties set by contract.

A large company may have a person for each. A practice with eight staff does not, and that is normal. In a small organization, the response is usually one or two people plus outside help:

  • Nora, the practice manager, decides on anything that stops the business and talks to the insurer and the authorities.
  • Sam is the first responder: he records, checks and acts on the technical side.
  • Bob, an external forensic consultant at example.org, can be engaged when the case needs more depth than Sam can give.
  • The portal provider holds logs that Northwind cannot see on its own, so its contact details must be ready.

Lesson 3 turns this into a plan with names, phone numbers and decisions written down in advance.

06How this course works

You will learn by doing, but always on safe ground. Three rules shape every lab:

  • Your own lab VM. From lesson 4 you work in an Ubuntu virtual machine that you build yourself, kept separate from your everyday computer.
  • Published practice images. Disk and memory images come from collections made for training: NIST CFReDS and Digital Corpora. They tell their own fictional stories, separate from Northwind's.
  • A case log. From lesson 3 you keep a notebook or text file, and every lab adds entries to it, just as Sam would.
Law and ethics

You may only investigate systems you own or are explicitly authorized, ideally in writing, to examine. That excludes a partner's phone, a friend's laptop and your employer's servers "just to help." In most countries, examining someone else's device or account without authority is illegal, even with good intentions. Everything in this course uses your own VM and data published for training.

Practical exercise: Sam's first ten minutes

Paper and pen, or a text file. Use the fictional Northwind Dental story from section 01. Check off each step as you finish it. Your checkmarks are saved in this browser.

Key takeaways

  • Events are everything observable, adverse events have a negative effect, and incidents threaten confidentiality, integrity, or availability, or break law or policy.
  • Incidents are not only outside attacks: insider mistakes and policy breaches count too.
  • Incident response aims to limit damage, restore operations, find the root cause, and improve.
  • Digital forensics follows four phases: collection, examination, analysis, reporting. What you collect is evidence, even without a court case.
  • Write the first observation down with date, time, and time zone before touching anything, and only ever examine systems you own or are authorized to examine.