Join our Newsletter — 33% off our NHI Course
Home› Guides› MITRE ATT&CK T1212 Exploitation for Credential Access: Detecting…
Guide Identity Visibility, Posture & Threat Detection

MITRE ATT&CK T1212 Exploitation for Credential Access: Detecting and Preventing Credential Theft

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 11 November 2026 11 min read
On this page

When the software that issues, checks or stores credentials has a bug, an adversary can mint tickets, forge tokens or read secrets out of memory without ever guessing a password. MITRE ATT&CK records this as T1212 Exploitation for Credential Access in the Credential Access tactic (ATT&CK v20, platforms Windows, Linux, macOS and Identity Provider). The bug classes themselves are explained in the Memory Corruption Vulnerabilities Guide; this page is about the credential consequence. It covers which components are in scope, how to prioritise their patches, the v20 detection analytics, the platform protections that limit what a successful exploit yields, and the response step that is most often missed: rotating every credential the vulnerable component could have exposed. It is for vulnerability managers, identity engineers and SOC analysts, and contains no exploit detail.

Key takeaways

  • Credential components are a patch class of their own. Domain controllers, Kerberos and Netlogon, identity providers, token services, VPN and SSO appliances, databases and secret stores deserve the shortest patch deadlines you have.
  • Patching does not un-leak a secret. If a flaw could expose credentials, rotate everything the component held or handled during the exposure window.
  • Validation logic is part of the attack surface. ATT&CK cites forged Kerberos tickets, replayed authentication and cloud token creation flaws; strict signature, issuer and audience checks remove whole families of bugs.
  • Platform isolation limits the prize. Credential Guard and LSA protection keep NTLM hashes and Kerberos TGTs out of reach of many endpoint exploits.
  • Detection looks for impossible credentials. DET0174 watches for anomalous ticket or token issuance, crashes of authentication services and unexpected LSASS access.

What T1212 covers in ATT&CK v20

ATT&CK describes adversaries exploiting software vulnerabilities to collect credentials or to circumvent authentication. Its examples are MS14-068, which let a domain user forge Kerberos tickets; replay attacks against services that do not properly validate authentication requests; and cloud flaws that allowed unintended token creation and renewal. It notes that exploitation for credential access may also achieve privilege escalation, depending on the process targeted.

AttributeATT&CK v20 valueDefender implication
TechniqueT1212 Exploitation for Credential AccessNo sub-techniques
TacticCredential AccessOften chained with T1068 Exploitation for Privilege Escalation
PlatformsWindows, Linux, macOS, Identity ProviderCloud identity services are in scope, not only on-premises software
Groups and campaignsUNC3886; Leviathan Australian IntrusionsSee the UNC3886 Threat Actor Profile Guide
MitigationsM1013 Application Developer Guidance; M1019 Threat Intelligence Program; M1048 Application Isolation and Sandboxing; M1050 Exploit Protection; M1051 Update SoftwareBuild, isolate, harden, patch and watch
Detection strategyDET0174 Detection Strategy for Exploitation for Credential AccessAnalytics AN0493 (Windows), AN0494 (Linux), AN0495 (macOS), AN0496 (Identity Provider)

How T1212 differs from its neighbours

Where credential-exposing flaws live

Component classWhat a flaw can yieldIllustrative caseContainment step after patching
Kerberos KDC and ticket validationForged tickets with elevated privilegesCVE-2014-6324 (MS14-068), KEV-listed, cited by ATT&CKReview privileged ticket issuance for the exposure window; consider resetting the KRBTGT account per your AD recovery plan
Machine account and secure channel protocolsTakeover of a domain controller's credentialsCVE-2020-1472 (Netlogon), KEV-listedConfirm enforcement mode; review machine account password changes
Memory disclosure in servicesPasswords, keys and tokens from process memoryThe MongoBleed breach, where an unauthenticated flaw returned heap memory that could hold credentialsRotate every credential the service held or handled
Identity provider and token servicesTokens issued or renewed without valid authenticationATT&CK cites a cloud flaw allowing unintended token creation and renewalRevoke sessions and refresh tokens; review issuance logs
Authentication protocols without replay protectionReuse of captured authentication messagesATT&CK replay attack examplesAdd nonces, timestamps or sequence numbers (M1013)
Endpoint credential stores (LSASS, keychains, PAM modules)Hashes, tickets and cached secretsDET0174 analytics for LSASS access and authentication daemon crashesReset credentials of every user who signed in to the host

Patch prioritisation for credential components

A vulnerability in a component that issues or stores credentials has a blast radius well beyond the host, so it should outrank a flaw of the same CVSS score elsewhere. NHI Mgmt Group recommends a simple rule set:

  1. Tag credential components in the asset inventory: domain controllers, identity providers and federation servers, PAM and secrets platforms, certificate authorities, VPN and SSO gateways, databases holding application credentials, and CI/CD runners that hold deployment keys.
  2. Treat KEV listing as a deadline: any CISA KEV entry affecting a tagged component goes to emergency change. See the CISA KEV Catalogue Guide and the Emergency Out-of-Band Patching Playbook.
  3. Read the advisory for credential impact: if it mentions memory disclosure, authentication bypass, token or ticket forgery, or key exposure, plan rotation alongside the patch.
  4. Use vendor enforcement phases: some fixes ship in audit then enforcement phases. For CVE-2020-1472, Microsoft's initial deployment phase started on 11 August 2020 and the enforcement phase on 9 February 2021, with Netlogon events 5827 to 5831 recording denied or allowed vulnerable connections.
  5. Isolate what cannot be patched (M1048): restrict network reach to the management and authentication ports of unpatchable components and monitor them closely.

Decision guidance

  • Patch now or patch and rotate: patch only when the flaw cannot expose secrets; patch and rotate when the advisory mentions memory, keys, tokens or authentication bypass, or when exploitation in the wild predates your patch.
  • Rotate everything or a subset: rotate everything the component held if it was internet-reachable while vulnerable; for internal-only exposure, rotate privileged and non-human credentials first and the rest on normal schedule.
  • Small organisations: rely on managed identity services and automatic updates for credential components, and keep a written list of secrets per service so rotation is possible at short notice.
  • Large organisations: integrate threat intelligence (M1019) so advisories for tagged components page the identity team directly, not only the patch team.

Platform protections that limit the yield

ProtectionWhat it protectsKey settings and limits
Credential GuardNTLM hashes, Kerberos TGTs and credentials stored by applications as domain credentials, isolated with virtualization-based securityEnabled by default on Windows 11 22H2 and Windows Server 2025 where requirements are met; does not protect the AD database or SAM; not recommended on domain controllers. See Credential Guard and VBS
LSA protectionLSASS memory against reads and code injection by non-protected processesRunAsPPL under HKLM\SYSTEM\CurrentControlSet\Control\Lsa (1 with UEFI lock, 2 without); audit with CodeIntegrity events 3065 and 3066; verify with WinInit event 12
Exploit protection (M1050)Memory corruption exploitation in generalProcess mitigations and control flow integrity; see the Exploit Mitigations Guide
Strict token validation (M1013)Forged or replayed tokensRFC 8725 practices: algorithm verification (3.1), validate all cryptographic operations (3.3), issuer and subject (3.8), audience (3.9), do not trust received claims (3.10); see the JWT Security Guide
Short-lived credentialsThe value of anything leakedPrefer workload identity and automatically rotated secrets for service-to-service access

Detection

PlatformATT&CK v20 analytic focus (DET0174)TelemetryTuning
Windows (AN0493)Forged Kerberos tickets, abnormal LSASS memory access, replayed authentication, crashes of authentication servicesSecurity Event IDs 4768, 4769 and 4770; Sysmon Event ID 10High-value accounts to watch; replay detection window
Linux (AN0494)Anomalous PAM authentications, segfaults in authentication daemons, exploitation followed by unauthorised loginsauditd execve records for processes touching sshd, gdm or login; network connection dataMonitored services; failure threshold
macOS (AN0495)opendirectoryd crashes, unauthorised Keychain API use, unusual sudo or login eventsUnified log; osquery process eventsWatched APIs; crash correlation window
Identity Provider (AN0496)Anomalous token creation or renewal, authentication bypass events, API abuse to mint tokens, gaps in audit logsSign-in logs for token issue and renewal; unified audit for consent grantsToken anomaly threshold; privileged app integrations
  • Tickets that should not exist: service tickets for privileged groups without a matching TGT request, or tickets with lifetimes or encryption types outside policy. See the Kerberos Security Guide.
  • Tokens without a sign-in: access tokens used for a user or application with no corresponding authentication event at the identity provider.
  • Use after disclosure: logins with database or service credentials from new sources after a disclosure date, which the MongoBleed page also recommends watching.
  • Service instability: repeated crashes of LSASS, sshd, opendirectoryd or an identity appliance process deserve a security look, not only a restart.

Metrics

  • Median time to patch KEV-listed flaws on tagged credential components.
  • Percentage of credential-exposure advisories closed with a completed rotation record.
  • Percentage of eligible Windows endpoints with Credential Guard and LSA protection running.
  • Number of credential components with no owner or no rotation runbook, target zero.
  • Percentage of token-validating services tested against RFC 8725 checks.

Response runbook

  1. Bound the window: establish when the component became vulnerable, when exploitation was possible and when the patch landed.
  2. Patch or isolate: apply the fix or cut network reach to the component.
  3. Enumerate exposed credentials: list every password, key, token, certificate and ticket the component stored, issued or processed during the window, including non-human ones.
  4. Rotate in order: signing keys and privileged secrets first, then service credentials, then user sessions. Revoke refresh tokens and sessions issued in the window.
  5. Hunt for use: search for authentications with the exposed credentials from unfamiliar sources, before and after rotation.
  6. Record and review: note which credentials could not be rotated quickly and fix that before the next advisory.

Where identity fits

T1212 is where vulnerability management and identity security meet. The credentials most often exposed by these flaws are non-human: database connection strings, API keys, service account passwords and signing keys that live in the memory or configuration of the vulnerable component. Organisations that know which secrets each service holds, and can rotate them automatically, turn a credential-exposure advisory into a routine change. See the Signing keys and certificates in breaches guide.

Framework mapping

FrameworkReferenceRelevance
NIST SP 800-53 Rev. 5SI-2 Flaw Remediation; RA-5 Vulnerability Monitoring and ScanningFinding and fixing credential component flaws
NIST SP 800-53 Rev. 5SI-16 Memory Protection; IA-2(8) Replay Resistant; SC-23 Session AuthenticityExploit mitigation, replay resistance, session integrity
NIST SP 800-53 Rev. 5IA-5 Authenticator Management; IR-4 Incident HandlingRotation after exposure
NIST CSF 2.0ID.RA-01; PR.PS-02; DE.CM-09; RS.MI-01Vulnerability identification, software maintenance, monitoring, containment

Common failure patterns

  • Patch and forget: the vulnerable service is updated, but the keys and passwords it exposed stay valid for months.
  • Credential components rated like any other server: a domain controller or identity appliance waits for the monthly cycle because its CVSS score was not the highest that month.
  • Enforcement phases left in audit: exemptions granted during a compatibility phase are never removed.
  • Home-grown token validation: services accept tokens without checking algorithm, issuer or audience, so a single signing or issuance bug has wide reach.
  • Crashes treated as availability issues: repeated authentication service crashes are restarted without anyone asking why.

Practitioner checklist

  • Tag every credential-issuing or credential-storing component in the asset inventory.
  • Apply emergency patch deadlines to KEV entries affecting those components.
  • Keep a list of secrets and keys per component so rotation can start within hours.
  • Run Credential Guard and LSA protection on eligible Windows endpoints and verify they are active.
  • Validate tokens per RFC 8725, including algorithm, issuer and audience checks.
  • Deploy DET0174-style detections for anomalous tickets, tokens and authentication service crashes.
  • Remove compatibility exemptions once enforcement phases are reached.
  • Include rotation and token revocation in every credential-exposure response.

Standards and references

Related NHI Mgmt Group resources: Memory Corruption Vulnerabilities Guide · CISA KEV Catalogue Guide · Credential Guard and VBS · JWT Security Guide · UNC3886 Threat Actor Profile Guide

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 11 November 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org