10 Active Directory Attacks Every SOC Analyst Must Know in 2026

Learn how SOC analysts detect and investigate password spraying, Kerberoasting, Pass-the-Hash, DCSync, Golden Tickets and other Active Directory attacks using Windows events, identity telemetry and structured incident-response techniques.

By Sanjay Verma, CISSP, CCSP, C|CISO | Published May 26, 2026 | Identity & Access Management (IAM) | 16 min read

10 Active Directory Attacks Every SOC Analyst Must Know in 2026 cybersecurity training article
Identity Threat Detection • 2026

10 Active Directory Attacks Every SOC Analyst Must Know

Learn how attackers target passwords, Kerberos tickets, privileged groups and domain controllers—and how a Tier-1 analyst can recognise the evidence, establish scope and escalate correctly.

KerberoastingPass-the-HashDCSyncGolden Ticket
Active Directory identity attack path A defensive illustration showing a suspicious login progressing toward credentials, privilege and domain control while a SOC analyst reviews security signals. IDENTITY ALERT Unusual ticket and privilege activity 1 2 3 4 5 ENTRYCREDENTIALS PRIVILEGELATERAL DOMAIN SOC DEFENCE MINDSET Correlate identity + endpoint + directory evidence
10 attacksBeginner-friendly explanations
15+ eventsWindows evidence to review
1 scenarioEnd-to-end SOC investigation
MITRE mappedCurrent defensive references

Active Directory is not merely a list of usernames and computers. In many enterprises, it is the control plane for authentication, privileged access, servers, applications and business data. If an attacker gains control of a highly privileged identity or the domain itself, a single endpoint compromise can become an organisation-wide incident.

For SOC analysts, the challenge is that identity attacks rarely appear as one obvious “Active Directory attack” alert. The evidence is distributed across domain-controller security events, endpoint telemetry, Kerberos activity, network connections, account changes and administrative actions. Strong analysts learn to correlate these signals instead of judging any event in isolation.

Start with the foundation If terms such as domain controller, forest, Kerberos, NTLM, LDAP, Group Policy or security group are unfamiliar, first read Active Directory for Cybersecurity: Complete Beginner Guide 2026.
Ethical and legal use This guide explains defensive detection and response concepts. Practise only in an isolated lab or an environment where you have explicit written authorisation. It intentionally avoids credential-theft commands and attack-execution instructions.

What You Will Learn

How common AD attacks workUnderstand the attacker’s objective without relying on offensive commands.
Which evidence mattersConnect Kerberos, logon, directory-change and endpoint signals.
How Tier-1 should respondValidate, scope, document and escalate without exceeding authority.
How defenders reduce riskApply least privilege, stronger authentication, auditing and secure administration.

Table of Contents

  1. Why AD attacks matter
  2. The identity attack chain
  3. Password spraying
  4. Kerberoasting
  5. AS-REP roasting
  6. Credential dumping
  7. Pass-the-Hash
  8. Pass-the-Ticket
  9. DCSync
  10. Golden Ticket
  11. Privileged-group manipulation
  12. Group Policy abuse
  13. Event ID detection table
  14. SOC investigation scenario
  15. Defensive checklist
  16. Frequently asked questions

Why Active Directory Attacks Matter

Attackers target identity because valid credentials can make malicious activity look similar to normal administration. A stolen password, NTLM hash or Kerberos ticket may allow an attacker to authenticate, access remote systems, add privileges or persist without exploiting another software vulnerability.

Microsoft’s current Active Directory security guidance recommends reducing attack surface, applying least privilege, protecting privileged accounts and using secure administrative hosts. Microsoft also warns that privileged groups possess rights capable of making extensive changes across AD and domain-joined systems.

Identity evidenceLogon results, Kerberos requests, NTLM usage, account status, MFA and source context.
Directory evidenceUser creation, group membership, object permissions, replication activity and GPO changes.
Endpoint evidenceProcesses, services, remote execution, credential access and unusual parent-child relationships.
Network evidenceConnections between workstations, servers and domain controllers, especially unusual lateral paths.

The Active Directory Identity Attack Chain

Not every incident follows the same sequence, but this simplified chain helps a Tier-1 analyst understand how one identity event may connect to a larger compromise:

Stage 1Initial AccessPhishing, exposed service or infected endpoint
Stage 2Credential AccessPassword, hash, token or Kerberos ticket
Stage 3PrivilegeAdmin identity, group change or delegated right
Stage 4Lateral MovementRemote access to systems and servers
Stage 5Domain ControlReplication, policy or persistence abuse
Tier-1 investigation rule Always ask what happened immediately before and after the identity alert. The meaningful story is often a sequence: suspicious process → credential activity → unusual login → privileged action.

The 10 Active Directory Attacks

1

Password Spraying

One common password tested against many accounts

MITRE T1110.003

Password spraying differs from a classic brute-force attack. Instead of trying many passwords against one account, the attacker tests a small number of commonly used passwords across many usernames. This approach may reduce account lockouts and make each individual account appear to have only a few failures.

Detection cluesOne source targets many accounts; similar failure reason; low failure count per user; activity may recur slowly.
Evidence to review462547714768 VPN, identity-provider and firewall logs.
Tier-1 responseValidate source and target count, identify any later success, check privileged accounts, establish scope and escalate confirmed compromise.

False-positive considerations: misconfigured services, stale stored credentials, vulnerability scanners and shared infrastructure can create repeated failures. The pattern across users, source devices, timing and subsequent activity is more useful than a single threshold.

2

Kerberoasting

Service-ticket material targeted for offline password cracking

MITRE T1558.003

Kerberoasting abuses a normal Kerberos capability. An authenticated user can request a service ticket for an account associated with a Service Principal Name. If the service account uses a weak password or legacy encryption, the attacker may attempt to crack the captured ticket material offline.

MITRE’s current Kerberoasting detection strategy recommends examining anomalous Event ID 4769 activity, unusual service-ticket volume, RC4 use and service accounts requested outside their normal baseline.

Detection cluesBurst of service-ticket requests, unusual requester, many different SPNs, legacy encryption or a rarely used service account.
Evidence to review4769 Requesting user, client address, service name, ticket encryption and endpoint processes.
Tier-1 responseConfirm whether requests fit normal application use, scope affected service accounts and escalate suspicious activity with ticket and endpoint evidence.
Do not over-alert on one fieldRC4 can be a useful risk signal, but it is not proof of Kerberoasting by itself. Correlate volume, requester behaviour, service account, endpoint telemetry and environmental baselines.
3

AS-REP Roasting

Accounts without Kerberos preauthentication are targeted

MITRE T1558.004

Kerberos preauthentication normally requires an account to prove knowledge of its credential before the Key Distribution Center returns authentication material. If preauthentication is disabled for an account, an attacker may request material that can be subjected to offline password guessing.

Detection cluesRequests involving accounts configured without preauthentication, unusual source host and concentrated discovery activity.
Evidence to review47684738 Account configuration, request source and related endpoint activity.
Tier-1 responseVerify whether the configuration is required, identify exposed accounts, look for successful logins and escalate unauthorised or unexplained changes.

Defensive priority: remove “Do not require Kerberos preauthentication” unless a documented legacy requirement exists, use strong managed credentials and monitor configuration changes.

4

Credential Dumping from Windows Systems

Credential material is extracted from memory or protected stores

MITRE T1003.001

After compromising a Windows device, an attacker may attempt to access credential material associated with LSASS or other operating-system stores. The objective may be plaintext credentials, NTLM hashes, Kerberos tickets or secrets that enable privilege escalation and lateral movement.

Detection cluesUnexpected process access to LSASS, suspicious memory access, security-tool tampering or unusual credential-related process behaviour.
Evidence to review46884672 EDR process tree, process access telemetry, file creation and user privileges.
Tier-1 responsePreserve alert details, identify process ancestry and user, determine whether credentials may be exposed, scope related hosts and escalate urgently.

Endpoint telemetry is often more valuable than domain-controller events for the original credential-access action. Correlate EDR with later authentication from the compromised host.

5

Pass-the-Hash

A stolen NTLM hash is used as authentication material

MITRE T1550.002

Pass-the-Hash allows an attacker to authenticate to compatible Windows services using a captured password hash rather than the original plaintext password. It is commonly associated with lateral movement after local or domain credentials have been exposed.

Detection cluesUnexpected NTLM network logons, privileged account use from an unusual workstation, remote administration and rapid host-to-host movement.
Evidence to review462446484672 Logon type, authentication package, source device and remote-service telemetry.
Tier-1 responseMap the source and destination systems, identify the account’s normal admin path, review preceding credential alerts and escalate suspicious lateral movement.

Context matters: NTLM is still present in many legitimate environments. A detection should consider source trust, account privilege, destination sensitivity, logon type, process activity and whether the sequence matches approved administration.

6

Pass-the-Ticket

A stolen Kerberos ticket is replayed for access

MITRE T1550.003

In Pass-the-Ticket activity, an attacker uses stolen Kerberos authentication material to access resources without knowing the account’s password. Depending on the ticket and account, this can support lateral movement or access to a specific service.

Detection cluesKerberos use from a device not normally associated with the account, unusual service access, identity and device mismatch or abnormal ticket sequence.
Evidence to review462447684769 Client address, service, ticket details and endpoint processes.
Tier-1 responseBuild the authentication timeline, compare normal device use, examine the originating endpoint and scope services accessed with the identity.

Ticket-based activity is difficult to judge from a single event. Combine domain-controller events, endpoint telemetry, identity baselines and resource-access logs.

7

DCSync

Directory replication rights are abused to request credential data

MITRE T1003.006

DCSync impersonates directory-replication behaviour. An attacker with powerful replication permissions may request password-related data from a domain controller without interactively logging on to the controller. Because legitimate domain controllers replicate, the critical question is whether the requesting identity and host should perform that activity.

Detection cluesReplication access requested by a non-domain-controller host, unusual privileged account or newly delegated replication right.
Evidence to review46625136 Directory-service access, object permissions, source host and privileged-account timeline.
Tier-1 responseTreat unexplained replication from a non-DC as high priority, validate identity and source, preserve evidence and escalate immediately.
Auditing prerequisiteEvent 4662 can be high volume and depends on audit and object-auditing configuration. A missing event does not prove that replication abuse did not occur.
8

Golden Ticket

Forged Kerberos TGTs create powerful domain persistence

MITRE T1558.001

A Golden Ticket is a forged Kerberos Ticket Granting Ticket created using compromised KRBTGT account material. It can allow an attacker to create authentication material representing highly privileged access and maintain persistence in the domain.

Detection cluesKerberos sessions with identity, privilege, device or ticket-lifetime anomalies; service access inconsistent with normal TGT issuance; known KRBTGT exposure.
Evidence to review4624467247684769 Cross-system timeline and identity consistency.
Tier-1 responseEscalate suspected ticket forgery as a major identity incident, scope privileged access and avoid making unsupported conclusions from one event.

Golden Ticket investigations are usually beyond Tier-1 closure authority. A senior incident-response and identity team may need to assess domain-wide exposure and recovery. Resetting KRBTGT is a controlled operational procedure, not a casual Tier-1 action.

9

Privileged-Group Manipulation

An account is added to a powerful security group

MITRE T1098

An attacker who compromises an account with directory permissions may add another identity to Domain Admins, Enterprise Admins, Administrators or another delegated group. The change may provide immediate privilege or establish persistence for later use.

Microsoft’s privileged accounts and groups guide explains that built-in privileged identities can perform extensive changes in Active Directory and on domain-joined systems.

Detection cluesUnexpected member added outside a change window, unusual creator account, temporary membership followed by sensitive action or nested-group change.
Evidence to review4728473247565136 Change ticket and subsequent logons.
Tier-1 responseVerify authorisation, identify who initiated the change, review actions by the added account and escalate if approval cannot be confirmed quickly.
10

Group Policy Abuse

A domain policy is modified to distribute malicious configuration

MITRE T1484.001

Group Policy can apply security settings, scripts, software and system configuration to many users or computers. If an attacker gains the ability to modify a high-impact Group Policy Object, the same central-management power can be abused for execution, persistence, weakened controls or large-scale disruption.

Detection cluesUnexpected GPO change, modified startup or logon script, new scheduled configuration, audit-policy reduction or SYSVOL change.
Evidence to review513647394719 Directory-change details, GPO history, SYSVOL and administrator endpoint activity.
Tier-1 responseIdentify the changed object and actor, validate change approval, determine linked users or devices and escalate suspicious high-impact modifications.

A GPO change is not malicious merely because it affects many systems. The analyst must compare the object, administrator, time, approved change and resulting configuration.

Windows Event IDs for Active Directory Attack Detection

Microsoft maintains an events-to-monitor reference and current advanced audit policy guidance. Your exact event availability depends on operating-system version, audit policy, object auditing, log collection and SIEM onboarding.

Event IDMeaningSOC use
4624Successful logonReview logon type, authentication package, source, destination and account.
4625Failed logonIdentify password spraying, brute force, stale credentials and source patterns.
4648Logon attempted with explicit credentialsCorrelate RunAs, scheduled tasks, administration and unusual credential use.
4672Special privileges assigned to a new logonValidate privileged sessions and correlate with the originating authentication.
4688New process createdReview suspicious tools, scripts and parent-child process relationships.
4719System audit policy changedDetect attempts to reduce visibility or unauthorised policy modification.
4720 / 4726User created / deletedInvestigate persistence accounts and suspicious identity lifecycle changes.
4728 / 4732 / 4756Member added to security groupMonitor global, local and universal group membership changes.
4738User account changedReview account flags, profile changes and unexpected configuration updates.
4768Kerberos TGT requestedInvestigate authentication source, account, encryption and AS-REP patterns.
4769Kerberos service ticket requestedDetect unusual service-ticket volume and possible Kerberoasting.
4771Kerberos preauthentication failedSupport password-spraying and authentication-failure analysis.
4662Operation performed on a directory objectInvestigate replication-right use and sensitive object access when auditing exists.
5136Directory service object modifiedReview group, permission, GPO and other directory-object changes.

For a wider event reference with beginner explanations, bookmark Windows Event IDs Every SOC Analyst Should Know.

Event IDs are evidence—not verdicts A normal administrator and an attacker may generate similar events. A verdict should consider the actor, source device, asset criticality, change approval, process activity, historical baseline and what happened next.

Real SOC Scenario: From Password Spray to Domain Escalation

Alert sequence

The SIEM identifies failed logins across 34 accounts from one VPN source. Twenty minutes later, one account succeeds. The same identity then requests many Kerberos service tickets and is added to a privileged operations group.

22:044625 and 4771 failures across many users
22:234624 success for one targeted identity
22:31Unusual 4769 service-ticket activity
22:424728 privileged-group membership change

How a strong Tier-1 analyst should respond

  1. Validate: confirm event times, data sources, alert rules and affected domain.
  2. Enrich: identify the VPN source, user, endpoint, privileged group and business owner.
  3. Build the timeline: correlate failed authentication, success, ticket activity and group change.
  4. Scope: search for the source across other accounts and for the user across endpoints, services and directory changes.
  5. Assess impact: identify whether privileged access was used after the group membership change.
  6. Escalate: provide a concise evidence summary and follow the identity-compromise playbook.
  7. Document: record queries, findings, timestamps, affected entities and actions taken.

Example escalation summary: “A single VPN source generated authentication failures across 34 accounts, followed by a successful login for one target. That identity then showed abnormal service-ticket requests and was added to a privileged group outside the approved change window. The sequence suggests credential compromise with possible privilege escalation. Scope currently includes one confirmed account, one VPN source and one privileged-group change; further endpoint and identity response is required.”

A Repeatable Tier-1 Investigation Method

  1. Confirm the signal: detection name, logic, source log, event time and data quality.
  2. Identify the entities: user, device, domain controller, source IP, service, group and object.
  3. Compare with baseline: normal source devices, administrative hours, authentication methods and service usage.
  4. Correlate telemetry: domain-controller events, EDR, VPN, firewall, DNS and identity-provider logs.
  5. Build before-and-after context: review activity before the first alert and after any successful authentication.
  6. Establish scope: search the same source, account, ticket, process or change across the environment.
  7. Decide and escalate: explain the evidence, uncertainty, affected assets, business risk and next action.

This workflow aligns naturally with the investigation skills in the SOC Analyst Tier-1 Interview Guide and the technique-mapping approach explained in the MITRE ATT&CK Beginner Guide.

Active Directory Defensive Checklist

Keep privileged-group membership minimal and regularly reviewed.
Use separate administrative and daily-use identities.
Protect administration through trusted, hardened workstations.
Require stronger authentication for privileged operations where supported.
Remove legacy protocols and encryption through a tested migration plan.
Use long, unique and managed passwords for service accounts.
Remove unnecessary Kerberos preauthentication exceptions.
Restrict and review directory-replication permissions.
Monitor privileged-group, GPO and audit-policy changes.
Forward critical domain-controller events to the SIEM.
Correlate identity and endpoint telemetry in detection rules.
Maintain tested identity-incident response and recovery procedures.

Microsoft’s guidance for least-privilege administrative models recommends delegating the permissions needed for defined roles instead of giving broad domain-wide authority.

How to Practise These Skills Safely

Create an isolated Windows lab with one domain controller, one Windows client and a SIEM or log collector. Focus on defensive visibility:

  • Enable and verify the required audit categories.
  • Create approved test users and security groups.
  • Generate normal and failed authentication events.
  • Make an authorised group-membership change and review the evidence.
  • Search Kerberos events and compare normal service-ticket behaviour.
  • Build one timeline and write a Tier-1 escalation note.
  • Create a detection rule for an unusual privileged-group addition.

Use the SOC Home Lab Beginner Guide to design a safe practice environment. If you want to strengthen investigation logic further, review Microsoft Defender XDR for SOC Analysts and PowerShell for SOC Analysts.

Common Detection Mistakes

MistakeWhy it failsBetter approach
Treating every failed login as an attackUsers and services regularly mistype or retain old passwords.Analyse source, targets, timing, failure reason and later success.
Calling every RC4 ticket KerberoastingLegacy applications may still use RC4.Correlate ticket volume, requester, SPNs, baseline and endpoint evidence.
Watching only domain controllersCredential access may begin on a workstation.Correlate DC, EDR, network, VPN and identity-provider telemetry.
Ignoring approved changesLegitimate administrators make sensitive changes.Validate actor, ticket, time, endpoint and exact object modification.
Closing after the first explanationA benign-looking event may sit inside a malicious sequence.Review activity before and after the signal and establish scope.
Using fixed thresholds everywhereEnvironment size and normal behaviour differ.Combine thresholds with baselines, rarity and asset criticality.

Frequently Asked Questions

Which Active Directory attack should a beginner learn first?

Start with password spraying and privileged-group changes because the concepts and Windows events are easier to understand. Then study Kerberos fundamentals before moving to Kerberoasting, Pass-the-Ticket and Golden Tickets.

Is Kerberoasting the same as stealing a password?

No. Kerberoasting requests service-ticket material that may be attacked offline. Whether the attacker recovers a usable credential depends on factors such as the service account password and encryption configuration.

What is the difference between Pass-the-Hash and Pass-the-Ticket?

Pass-the-Hash uses a stolen password hash with compatible authentication, commonly NTLM. Pass-the-Ticket uses stolen Kerberos ticket material. Both can enable access without knowing the user’s plaintext password.

Why is DCSync considered high risk?

DCSync abuses directory-replication rights to request credential-related data. Unexplained replication activity from a non-domain-controller system can indicate severe domain compromise and requires urgent escalation.

Can one Windows Event ID confirm an AD attack?

Usually not. Event IDs describe actions, and legitimate administration may generate similar records. Analysts should correlate actor, source, device, privilege, process, timing, change approval and follow-on activity.

Which logs are most important for AD monitoring?

Domain-controller security events are essential, but mature monitoring also includes EDR, PowerShell, DNS, firewall, VPN, identity-provider, server and application logs.

Do all organisations generate the same AD events?

No. Event availability and fields depend on Windows version, audit policy, object auditing, log retention, collection agents and SIEM configuration. Validate your own telemetry before writing detections.

Should Tier-1 analysts reset KRBTGT after a Golden Ticket alert?

No. KRBTGT recovery is a controlled, high-impact identity operation that must follow the organisation’s incident-response and Active Directory recovery procedures. Tier-1 should preserve evidence and escalate.

Can I learn AD detection without offensive tools?

Yes. Beginners can learn authentication, group changes, auditing, SIEM queries, timeline building and incident writing through authorised administrative actions and simulated log datasets.

How can I demonstrate Active Directory skills in an interview?

Explain a small lab, the events you collected, one detection you created, the evidence you reviewed and how you wrote an escalation note. Demonstrable investigation reasoning is stronger than listing attack names alone.

Conclusion: Follow the Identity Story

Active Directory attack detection is not about memorising ten dramatic names. It is about recognising how identities, devices, tickets, privileges and directory changes connect.

A strong SOC analyst asks:

  • Who performed the action?
  • From which device and network source?
  • Was the identity expected to perform it?
  • What happened before and after?
  • Which users, services and systems are affected?
  • Does the evidence require urgent containment or escalation?

Practise answering these questions with real Windows events in an authorised lab. That is how attack terminology becomes job-ready investigation skill.

Build Practical Active Directory and SOC Investigation Skills

CybersecurityTRAIN.com helps freshers and IT professionals learn SIEM monitoring, Windows events, identity attacks, phishing investigation, incident response, MITRE ATT&CK and Tier-1 escalation through practical, role-focused training.

WhatsApp or call: +91 98857 89887  |  Email: trainings@thecyberseal.com

Training and career support can improve skills and interview readiness but cannot guarantee employment, promotion or salary outcomes.

About the Author

Sanjay Verma, CISSP, CCSP, C|CISO is a cybersecurity professional and mentor with extensive experience across enterprise security, cloud security, Zero Trust, managed security services, security architecture, risk management and career-focused cybersecurity training.

CybersecurityTRAIN.com is a cybersecurity training initiative and a unit of Cyberseal Infosec Solutions Pvt. Ltd.

Related articles