0%
Security Engineering for Software Teams
Thinking About Security
Cryptography You Will Actually Use
Identity
Vulnerability Classes
Infrastructure and Cloud
Incident Response and Forensics
Every organisation eventually faces a security incident. The difference between a contained event and a disaster is usually preparation: clear roles, practised procedures, ready tools and good relationships established before anything goes wrong. This section introduces the incident response lifecycle and the preparation that makes it work under pressure.
What counts as an incident
A security incident is an event that actually or potentially compromises the confidentiality, integrity or availability of systems or data, or violates security policy. Examples include a compromised account, malware on a laptop, exposed customer data, ransomware or an insider misusing access. Defining incidents clearly helps people know when to raise the alarm rather than quietly fixing something and moving on.
Events, alerts and incidents
Not every alert is an incident. An event is any observable occurrence; an alert is an event flagged for attention; an incident is a confirmed or strongly suspected security problem that requires a coordinated response. Triage turns alerts into incidents or closes them. Declaring an incident starts the formal process, with roles, communication and documentation.
The incident response lifecycle
Most frameworks describe a similar lifecycle. Preparation builds capability before incidents. Detection and analysis confirm what is happening. Containment limits damage. Eradication removes the attacker and their footholds. Recovery restores normal operations safely. Post incident activity captures lessons and drives improvements. In practice, phases overlap and loop as new information appears.
The incident response plan
An incident response plan documents how the organisation handles incidents: definitions, severity levels, roles, escalation paths, communication rules, legal and regulatory obligations and links to playbooks. A good plan is short enough to use under stress, approved by leadership and known to the people who will use it. Plans that live in a forgotten folder do little good.
Roles
Clear roles prevent confusion. The incident commander coordinates the response and makes decisions. Technical leads investigate and remediate. A scribe records actions and decisions with timestamps. A communications lead handles internal and external messages. Legal, privacy, executive and customer facing representatives join as needed. One person may hold several roles in a small incident, but each role should be explicitly assigned.
The incident commander
The incident commander does not do the technical work. They keep the big picture, set priorities, assign tasks, track progress, manage handoffs and decide when to escalate. This separation prevents the common failure where the most skilled engineer becomes absorbed in investigation and nobody coordinates. Commanders are trained in advance and rotate to avoid burnout.
Severity levels
Severity levels define how serious an incident is and what response it needs. Criteria usually include the sensitivity of affected data, the number of affected users or systems, operational impact, legal implications and whether the attacker is still active. Each level specifies who must be notified, how quickly and whether executives or external parties need to be involved.
Playbooks
Playbooks provide step by step guidance for common incident types: compromised account, malware infection, cloud key exposure, ransomware, data leak or denial of service. They list initial triage steps, evidence to collect, containment options, people to involve and communication templates. Playbooks speed response and reduce mistakes, but responders must adapt them to the specific situation.
Tools and access
Responders need tools ready before an incident: log search, endpoint investigation and isolation, cloud console access with appropriate roles, forensic imaging tools, secure communication channels and case management. Access should be tested regularly, including emergency accounts that work when normal identity systems are compromised. Discovering missing access during an incident costs critical hours.
Out of band communication
If attackers control email or chat, using those channels to coordinate the response can alert them. Preparing an out of band channel, such as a separate messaging workspace or phone bridge with pre distributed contact details, allows secure coordination. Teams should know when to switch to it, such as when identity or communication systems may be compromised.
Legal and regulatory preparation
Many incidents trigger legal obligations: notifying regulators, customers or individuals within defined timeframes, preserving evidence and handling law enforcement requests. Involving legal counsel in planning, knowing which laws and contracts apply and preparing notification templates avoids rushed decisions. Legal involvement can also affect how investigation findings are documented and shared.
External partners
Organisations often rely on outside help: incident response firms, forensic specialists, cyber insurance providers, cloud providers and law enforcement. Arranging retainers and contacts in advance means help arrives quickly. Insurance policies may require using approved firms or notifying the insurer promptly, so these terms should be known before an incident.
Asset and contact inventories
Responders need to know what systems exist, who owns them, how critical they are and how to reach owners at any hour. Up to date inventories and on call rotations shorten the time spent finding the right people. During a major incident, a shared contact list for executives, legal, communications and key vendors saves time.
Exercises
Tabletop exercises walk participants through a realistic scenario, testing decisions, communication and plan gaps without touching systems. Technical exercises test tools and procedures hands on. Regular exercises, including executives for major scenarios, build familiarity and reveal problems. Each exercise should produce actions that are tracked to completion.
Training responders
Responders improve through training on investigation techniques, the environment, playbooks and communication. Shadowing experienced responders during real incidents, rotating engineers through on call and reviewing past incidents build skills. Training non security staff to report suspicious activity quickly matters too, because early reports often come from them.
Measuring readiness
Readiness can be measured: percentage of critical systems with logging and endpoint coverage, time to obtain emergency access, playbook coverage of likely incident types, exercise frequency and completion of exercise actions. These measures show whether preparation is real or only on paper.
Culture of reporting
People must feel safe reporting mistakes and suspicious activity. If someone who clicked a phishing link fears punishment, they may stay silent while an attacker gains a foothold. A blameless culture, simple reporting channels and thanking reporters encourage fast reporting, which is one of the most effective ways to shorten incidents.
Preparation pays off
Teams that prepare respond faster, make fewer mistakes and recover more completely. Time spent on plans, playbooks, access, contacts and exercises during calm periods is repaid many times when a real incident arrives. Preparation also reduces stress for responders, who can follow practised routines rather than improvising every decision.
Break glass accounts
Break glass accounts provide emergency administrative access when normal identity systems fail or are compromised. They use strong, separately stored credentials, are excluded from risky dependencies such as the main identity provider where appropriate and trigger alerts whenever used. Testing them on a schedule confirms they still work, because an emergency account that fails during an incident is worse than useless.
Case management
A case management system holds the incident record: timeline, evidence references, tasks, decisions and communications. It replaces scattered chat threads and personal notes with a single source of truth. Restricting access to sensitive cases protects confidential details, while templates for each incident type prompt responders to capture the information that reviews and regulators will later need.
The response lifecycle runs from preparation through detection, containment, eradication and recovery to post incident review. Define severity levels and roles, especially an incident commander who coordinates rather than investigates, and prepare playbooks, tools, emergency access, out of band channels, legal guidance, partners and exercises.