What Is DFIR? Digital Forensics and Incident Response
Skip to content

What Is DFIR?

DFIR stands for digital forensics and incident response. It is the field of cybersecurity that investigates security incidents from evidence and manages the response to them.

Digital forensics establishes what happened. Incident response contains the threat and restores operations. This guide explains how the two fit together, how long each kind of evidence lasts, and the terms the field uses.

Forensics and Response Compared

DFIR joins disciplines that grew up separately and pull in different directions.

Digital forensicsIncident response
GoalEstablish the facts.Limit the damage.
QuestionWhat happened, when, and how do we know?What do we do now?
PaceCareful. Evidence must hold up later.Fast. The attacker may still be active.
Works onCopies and preserved records.Live systems.
ProducesA timeline, findings, and an evidence trail.Containment, eradication, and recovery.
Risk when done aloneAn accurate report about damage that could have been limited.Action on guesses, with the attacker left in place.

The disciplines are combined because each needs the other. Responders cannot contain what has not been found, and investigators lose their evidence when responders rebuild systems before it is collected.

In practice the tension is about time. Investigators want to observe and collect before anything changes. The business wants systems back. The answer is fast, targeted collection, described under One System or the Whole Fleet below, so that containment does not wait for a full analysis.

The Origins of DFIR

The halves of DFIR have separate histories, and the differences in outlook come from them.

Digital forensics grew out of law enforcement. As computers appeared in criminal cases in the 1980s and 1990s, investigators needed ways to recover evidence from them that would stand up in court. The rules that developed were procedural. Work on a copy, never the original. Document every step. Be able to explain to a judge exactly how a conclusion was reached.

Incident response grew out of operations. In November 1988 a self-spreading program known as the Morris worm disrupted a large part of the early internet, and the response was improvised by administrators phoning each other. Within weeks the first computer emergency response team was established at Carnegie Mellon University to coordinate future incidents. The aim was to understand enough to stop the damage, restore service, and prevent a repeat.

For years the two were done by different people. Forensic examiners worked in laboratories on seized equipment. Incident responders worked in operations centers on live networks.

They merged as intrusions changed. Attackers stopped causing immediate, visible damage and began staying inside networks for weeks, using valid accounts and ordinary tools. Stopping them required finding out precisely what they had done, which is a forensic question, while they were still active, which is a response problem. An organization could no longer do one and then the other.

How Long Evidence Lasts

Each kind of evidence has a lifetime, and much of the skill in DFIR is collecting in the right order.

EvidenceWhat it showsTypical lifetime
Memory of a running systemPrograms running now, including malware that never touched the disk, plus credentials and encryption keys.Until the system is restarted or powered off.
Active network connectionsWhat the system is talking to at this moment.Seconds to minutes.
Windows event logsLogons, new services, scheduled tasks, account changes.Hours to weeks. Logs have a fixed size and overwrite the oldest entries, and a busy domain controller fills its Security log fastest.
Firewall and VPN logs on the deviceRemote access and outbound connections.Days, unless sent to central logging.
Endpoint tool telemetryProcess, file, and network activity.Weeks to months, set by the product and license.
Cloud sign-in logsWho signed in, from where.7 to 30 days by default in Microsoft Entra ID.
Cloud audit logsMail, file, and administrative activity.180 days by default in Microsoft 365.
Volume shadow copiesEarlier versions of files and of the registry.Until deleted, which ransomware operators do before encrypting.
Disk artifactsEvidence that programs ran, files were opened, and devices connected.Months to years, until overwritten or the system is rebuilt.
BackupsThe state of systems on a past date.As long as the retention policy keeps them.

The table gives the order of work. Collect the short-lived evidence first, which is known as the order of volatility. Never switch off a system that may hold evidence in memory unless there is no other way to stop the damage.

The table also explains why many investigations cannot reach the start of an intrusion. If the attacker arrived six weeks ago and the firewall keeps five days, the first connection is gone.

How an Examiner Proves Something Happened

Forensic conclusions rest on corroboration. A single record can be wrong, missing, or planted. Several independent records that agree are hard to dispute.

Take one common question: did a particular program run on this computer, and when? A Windows system records that fact in several unrelated places.

The program execution cache

Windows workstations keep a record of programs that have run, with a count and recent run times, to speed up later launches. The feature is off by default on servers.

The application compatibility records

A separate mechanism notes programs the system has seen, with file details.

The user’s activity records

Per-user records show which programs that person started from the desktop.

The process creation log

Where auditing is enabled, each program start is logged with its command line and the account that launched it.

Endpoint security telemetry

The security agent keeps its own record of the process, its parent, and its network connections.

The file system

The program file carries creation and modification times, and the file system’s change journal keeps entries even if the file was later deleted.

Memory

If the program is still running, it can be seen directly.

An attacker who deletes the program removes one of these. One who clears the event log removes another, and the clearing itself is logged. Removing all of them without a trace is extremely difficult, and gaps where records should be are evidence too.

The examiner lines these up on a timeline. When five sources agree that a program first ran at 02:14 under a service account, the finding is solid. When they disagree, the disagreement usually reveals something: a changed clock, a tampered timestamp, or a second copy of the program.

This is why forensic reports read cautiously. A good one says what the evidence shows, how many sources support it, and what could not be checked.

Branches of Forensics

Endpoint forensics

Processes, files, registry entries, scheduled tasks, and user activity on workstations and servers.

Memory forensics

The contents of a running system’s memory.

Log analysis

Authentication, application, and security logs from systems and identity providers.

Network forensics

Connection records and captured traffic showing what communicated with what.

Cloud forensics

Audit and access records from cloud platforms and software services.

Email forensics

Message headers, mailbox access records, and rules.

Malware analysis

Examining malicious files to learn what they do and what they contact.

Mobile forensics

Evidence from phones and tablets, where relevant.

Each source answers a different question. Endpoint evidence shows what ran. Identity logs show who signed in. Network records show what left. An investigation that relies on one source sees one part of the picture.

Our guide to endpoint forensic artifacts covers what to collect on Windows, Linux, and macOS.

An Investigation From Start to Finish

1
TriageConfirm that something happened and how serious it looks. Output: a decision to declare an incident, and a first list of affected systems.
2
PreservationCapture memory and short-lived records, and stop logs being overwritten. Output: evidence that can no longer be lost.
3
ScopingSearch every system for the traces found so far. Output: the full list of affected accounts and systems, which is always longer than the first one.
4
ContainmentCut the attacker’s access across everything in scope at once. Output: an attacker who can no longer act.
5
AnalysisReconstruct the sequence: entry point, movement, what was accessed, what was taken. Output: a timeline with evidence for each entry.
6
Eradication and recoveryRemove persistence, reset credentials, rebuild and restore. Output: systems back in service and validated as clean.
7
ReportingState what happened and what should change. Output: a report that separates confirmed findings from open questions.

Steps 3 to 6 overlap. Scoping continues during containment, and analysis continues during recovery. The steps are in order of when each must start, not when each finishes.

The People in a DFIR Team

A serious incident needs several kinds of expertise. In a small team one person covers more than one role. In a large response each is a separate person or group.

Incident lead

Runs the response. Sets priorities, coordinates the technical work with the business, and keeps one current account of the facts.

Forensic analyst

Examines endpoints, memory, and logs to reconstruct what happened. Produces the timeline.

Threat hunter

Searches the wider environment for the same attacker elsewhere. Turns each new finding into a search across every system.

Malware analyst

Takes apart suspicious files to learn what they do, what they contact, and how to detect them.

Cloud and identity specialist

Investigates sign-ins, tokens, mailbox access, and cloud configuration. Increasingly central, since many intrusions never touch a traditional server.

Network analyst

Works from traffic records to show movement between systems and data leaving.

Recovery lead

Usually from IT operations. Plans and carries out the rebuild and restore, guided by what the investigation has cleared as safe.

Liaison roles

Legal counsel, privacy, communications, and the insurer’s representative, who turn technical findings into decisions and statements.

DFIR and the Law

Forensic work has legal consequences, which shape how it is done.

Admissibility. If evidence may be used in court, an employment dispute, or an insurance claim, it must be shown to be authentic and unaltered. This is the purpose of forensic images, hash values, and chain of custody records. Evidence gathered casually may be accurate and still be challenged.

Legal privilege. In many jurisdictions, an investigation directed by lawyers for the purpose of legal advice can be protected from disclosure. Organizations often engage a forensic provider through legal counsel for this reason. The rules vary and are easy to get wrong, so counsel should be involved early.

Notification duties. Forensic findings decide whether data breach and incident reporting obligations apply. The investigation has to answer specific questions, such as what personal information was accessed, within legal timeframes.

Expert evidence. A forensic examiner may be asked to give a report or testimony explaining findings to a non-technical audience. Clear documentation made at the time is what makes that possible months later.

Privacy and employment law. Examining an employee’s device or mailbox raises its own legal questions. Investigations into staff conduct should involve HR and legal counsel from the beginning.

DFIR Terms

Artifact

A record left by an operating system or application that shows something happened, such as a file that proves a program ran.

Chain of custody

The documented history of who collected a piece of evidence and who has handled it since.

Containment

Action that stops the attacker from doing further harm.

Dwell time

How long the attacker was present before detection.

Eradication

Removing the attacker’s tools, accounts, and access.

Forensic image

An exact copy of a disk or memory, taken so the original is not altered by analysis.

Hash

A fingerprint calculated from a file. If the file changes, the hash changes, so a matching hash shows the evidence has not been altered.

Indicator of compromise

A specific trace of the attacker, such as an address, a file name, or a hash, that can be searched for elsewhere.

Lateral movement

The attacker moving from one system to another inside the network.

Persistence

Any method the attacker uses to keep access after a restart or a password change.

Root cause

The weakness that allowed the incident to begin.

Timeline

The ordered sequence of events, each tied to its evidence.

Triage collection

A fast, targeted collection of the most useful evidence from a system, in place of a full disk image.

Volatile data

Evidence that is lost when a system is powered off.

One System or the Whole Fleet

Traditional forensics examines one system in depth, usually from a full disk image. It is thorough and slow. One system can take a day or more.

Intrusions rarely involve one system. An attacker with administrator access may have touched hundreds. Examining them one at a time is impossible in the time available.

Enterprise-scale DFIR works the other way round. It collects targeted evidence from every endpoint at once, then searches the results for the attacker’s traces. Systems that match are examined in depth.

Both approaches are used in the same investigation. Fleet-wide collection finds where the attacker has been. Deep analysis of selected systems explains exactly what happened there.

Who Does DFIR

Few organizations keep a full DFIR team. The skills are built through exposure: a person learns to investigate ransomware by investigating ransomware, and most organizations, fortunately, do not provide enough practice.

ApproachSuitsConsider
In-house teamLarge organizations with frequent incidents and sensitive environments.Cost, tooling, and keeping skills current.
Retained providerMost mid-sized and many large organizations.Response time, onboarding, and knowledge of the environment.
HybridOrganizations whose security team handles routine incidents.A clear handover point for major incidents.
Insurer-appointedOrganizations relying on their cyber policy.Limited choice, and no prior knowledge of the environment.

A security operations center and a DFIR team do different jobs. The SOC monitors continuously and handles alerts. DFIR investigates confirmed incidents in depth. Many organizations use a SOC day to day and bring in DFIR specialists when something serious is found.

When to Call a DFIR Team

  • The scope of an incident is unknown.
  • An attacker may still be active.
  • A privileged account or domain controller is involved.
  • Ransomware has been deployed, or data theft is suspected.
  • Evidence may be needed for legal, regulatory, or insurance purposes.
  • The internal team cannot collect evidence across all systems.

Call before systems are wiped or rebuilt.

DFIR With ThreatDefence

ThreatDefence provides digital forensics and incident response 24x7. Our responders bring their own platform to an engagement, so evidence collection across endpoints, identity, network, and cloud starts in the first hours, and our SOC watches the environment while the organization recovers.

Our forensic work covers memory, disk, log, network, email, and cloud evidence, collected under court-admissible procedures with chain of custody recorded. An incident response retainer gives priority access to the DFIR team.

Forensic collection in the ThreatDefence portal: artifacts collected from a host, evidence by source, collection progress, fleet hunt results and findings

Experiencing a breach?

Our incident response team answers 24x7, with or without a retainer.

Talk to Our Experts

Discuss digital forensics, incident response retainers, and 24x7 monitoring for your organization.

    I'm an MSP, MSSP, or reseller

    By submitting this form, you accept the terms in our privacy policy.

    Frequently Asked Questions

    What does DFIR stand for?

    DFIR is short for digital forensics and incident response. Forensics works out what happened from evidence, and incident response contains the attacker and restores service. The term is used for the discipline, for the teams that practice it, and for the service organizations retain so that help is on hand when an incident is declared.

    What is the difference between digital forensics and incident response?

    Forensics asks what happened and how we know. Incident response asks what to do now. They differ most in pace: a forensic examiner wants nothing changed until it is copied, and a responder wants the attacker out today. Done together, the first hours of collection give responders enough to contain the attacker without destroying what the examiner needs.

    What does a DFIR analyst do?

    A DFIR analyst collects and examines evidence from endpoints, logs, networks, and cloud services to reconstruct an incident. Day to day that means building a timeline, turning each finding into a search across the other systems, advising on containment, and writing down what the evidence supports and what it leaves open.

    Is DFIR the same as a SOC?

    No. A SOC watches continuously and decides which alerts are real. DFIR starts where that ends, with a confirmed incident that needs scoping, evidence, and a timeline. The handover works best when it is agreed in advance: what the SOC may contain on its own authority, when it calls DFIR, and what evidence it preserves before it acts.

    How long does a DFIR investigation take?

    A single compromised account may take a day or two, and a network-wide intrusion can take weeks. The length is set mostly by how many systems are in scope and how much evidence survived. Early findings guide containment long before the final report is written, so the business does not wait for the end of the investigation to act.

    What is chain of custody in digital forensics?

    Chain of custody is the documented record of who collected evidence, when and how it was collected, and who has handled it since. It matters outside court as well, because an insurer, a regulator, or the other side in an employment dispute can challenge evidence whose handling is undocumented. Starting the record at collection takes minutes, and it cannot be reconstructed afterward.

    What tools does a DFIR team use?

    By function: a remote collection agent that gathers evidence from many endpoints at once, memory and disk analysis tools, a log search platform, network traffic records, a malware sandbox, and a timeline that joins the results. Open-source tools cover most of these functions. What decides the outcome is whether the tooling can reach every system quickly, including servers and cloud services.

    Can DFIR be done remotely?

    Yes. Most investigation and containment is now performed remotely using endpoint agents, authorized cloud access, and remote evidence collection tools. On-site work may still be necessary for physical evidence acquisition, inaccessible systems, infrastructure recovery, and incident coordination. Remote work also starts sooner, which matters because the shortest-lived evidence is gone within hours.