Join our Newsletter — 33% off our NHI Course
Home› Guides› Help Desk Social Engineering Incident Playbook: What to…
Template Incident Response & Forensics

Help Desk Social Engineering Incident Playbook: What to do when a caller fooled your service desk

PDF for newsletter subscribers · free to read online

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 27 October 2026 10 min read
On this page

When a caller talks your service desk into resetting a password or MFA factor, the attacker leaves the call holding a fully valid identity that most controls will trust. CISA's advisory on Scattered Spider describes exactly this route: actors gather personal details and security question answers, then phone the help desk to reset passwords and MFA tokens, which gives them single sign-on to everything the victim can reach. The response is different from a phishing case because the "compromise" happened inside your own process, so you have to unwind every change the agent made, revoke every session and device that followed, and find out whether the same caller succeeded elsewhere. This template is for service desk managers, SOC leads and identity teams. Prevention and verification design sit in the Account Recovery and Help Desk Security Guide; this page is the response.

How to use this template

  • Adopt it with your tooling: replace [placeholders] with your identity provider (IdP), ITSM platform, telephony system and named people. Keep the commands and console paths for your IdP in an appendix the desk can reach offline.
  • Give the desk a trigger, not a burden: agents must be able to raise a suspected social engineering call in one step, without fear of blame. Speed of that first report decides how much damage follows.
  • Treat it as an identity incident first: the immediate actions are revoke, re-verify and review. Escalate to the Identity Provider Tenant Compromise Playbook if an administrator account was reset, and to the Active Directory Domain Compromise Response Playbook if domain credentials were taken.
  • Log decisions: every Decision point needs an owner and a timestamp in the incident record.

1. Scope and triggers

This playbook applies to [Organisation]'s internal IT service desk, outsourced desks acting for [Organisation], and any team that can reset credentials or MFA (identity operations, regional IT, privileged access team). Invoke it when any of the following occurs:

  • Agent suspicion: an agent reports a caller who pushed for urgency, failed or avoided verification, called from an unexpected number, or asked for an MFA reset for a privileged user.
  • User report: a user receives a password change or "new MFA method registered" notification they did not request, or is locked out after a reset they did not ask for.
  • Detection: a password or MFA reset followed within minutes by sign-in from a new device, anonymising proxy or residential VPN range, or by new MFA registration (ATT&CK T1556.006) or device registration (T1098.005).
  • Pattern: several calls in a short period naming different users, especially administrators or staff with help desk, finance or IdP roles.
  • Threat intelligence: a sector alert, such as the NCSC's May 2025 recommendations after the UK retail incidents, describing active help desk impersonation.

Voice phishing of end users by fake IT staff (T1566.004) is related; use this playbook when the desk itself acted on the caller's request, and the Phishing Incident Response Playbook otherwise.

2. Roles and contacts

RoleResponsibilitiesNamed lead / deputy
Incident lead (SOC)Triage, scoping, containment decisions, incident record[Name] / [Name]
Service desk managerPulls call recordings and tickets, identifies agents involved, enforces the heightened verification mode[Name]
Identity engineerSession revocation, MFA method review, IdP audit log queries, policy changes[Name]
Outsourced desk contactEvidence from the provider, their agent actions, their other clients if the caller reused tactics[Provider, 24x7 number]
HR and line managersVerify affected users through a known channel; support agents without blame[Name]
Legal, privacy, communicationsNotification assessment, staff messaging[Name]

Out-of-band verification list: keep a list, outside the IdP, of how to reach each privileged user and their manager by a known number. You will need it when you cannot trust the caller ID or the user's mailbox.

3. Detection and triage

  1. Freeze the ticket and the call (target: [15] minutes). Preserve the ticket, call recording, caller number, call time, agent ID and every action the agent took. Do not close or edit the ticket.
  2. Confirm with the real user through a known channel (manager call, in-person, video with ID). Never call back the number the caller gave.
  3. List what was changed. Password reset, MFA methods removed or added, Temporary Access Pass or bypass code issued, phone number changed, group or role assigned, device enrolled. Use IdP audit logs, not the ticket alone.
  4. List what happened next. Sign-ins after the reset (IP, ASN, device, user agent), new MFA registrations, application access, OAuth consents, mailbox rules, file downloads, and remote access tool installs.
  5. Hunt for sibling calls. Search the ITSM and telephony records for the same number, the same voice script, or resets for other users in the last [30] days.
  6. Classify. Use the table below.
SituationSeverityPath
Caller refused or failed verification, no change madeLowRecord, alert desk, monitor targeted user
Reset made, no sign-in by attackerMediumRevert, re-verify user, hunt for siblings
Reset made and attacker signed in, standard userHighFull containment, access review, data exposure scoping
Reset made for administrator, IdP or help desk role holderCriticalFull containment plus tenant compromise playbook

Where to look

SourceEvents or fields
Okta System Loguser.mfa.factor.reset_all, system.mfa.factor.* (activate, deactivate), system.idp.lifecycle.create, user.authentication.auth_via_IDP, user.session.start with securityContext.isProxy, user.account.report_suspicious_activity_by_enduser
Microsoft Entra IDAudit log password resets by admin and authentication method registration or deletion; sign-in logs for the user after the reset; Entra ID Protection risky sign-ins, including the Microsoft Entra threat intelligence detection the NCSC highlighted
On-premises AD4724 (password reset attempt), 4738 (account changed), 4728 / 4732 / 4756 (group adds)
ITSM and telephonyTicket history, agent actions, call recordings, caller numbers, call duration
EDRNew remote access tools (T1219) on devices the user signed in to

4. Containment

  1. Disable or block sign-in for the affected account while you re-verify the user. For hybrid accounts, disable in AD and in the IdP.
  2. Revoke sessions and refresh tokens. In Entra ID use Revoke sessions or Revoke-MgUserSignInSession; in Okta clear user sessions. Microsoft notes existing access tokens can last about an hour unless the app supports Continuous Access Evaluation, and that application session cookies are not revoked by Entra ID, so also sign the user out in high value SaaS apps directly. See the Token and Session Security Guide.
  3. Remove attacker MFA methods and devices. Delete every authenticator, phone number and device registered since the reset; disable registered devices.
  4. Reset the password twice for on-premises accounts, as Microsoft advises when pass-the-hash is possible, then hold the new credential until the real user is verified.
  5. Raise the desk to heightened verification mode: no MFA resets for privileged users without [two-person approval / in-person or video ID check]; all resets for named target groups go to [identity team]. Okta's guidance recommends constraining help desk roles so they cannot reset highly privileged administrators.
  6. Block the infrastructure seen: attacker IPs, proxy ranges, rogue remote access tools and any rogue identity provider or federated domain added.

Decision: if more than [three] users are affected, or any administrator, the incident lead decides whether to suspend all remote MFA resets for a period. The trade-off is user lockouts against a live attacker using your desk as a tool. Record the decision and the time it ends.

5. Eradication

  • Undo every change: group and role assignments, OAuth consents, mailbox forwarding rules, new accounts (T1136), and any policy changes such as removed MFA requirements.
  • Check IdP configuration: new identity providers, inbound federation, federated domains and custom admin roles. Okta's 2023 investigation found attackers adding a second IdP to impersonate users; Microsoft lists "Suspicious domain added to Microsoft Entra ID" as an Octo Tempest detection.
  • Remove persistence on endpoints: remote access tools, tunnelling utilities and scheduled tasks on any device the attacker reached.
  • Rotate exposed secrets: credentials the user could read in vaults, wikis, chat or code repositories, including API keys and service account passwords. See the Infostealer Logs Monitoring and Response Guide for credential rotation at scale.
  • Close the process gap that was used: for example, knowledge based questions that the attacker answered from public data.

6. Recovery

  1. Re-verify the real user using the strongest method available: in person with ID, verified video with manager present, or an existing phishing-resistant authenticator.
  2. Re-enrol with phishing-resistant MFA where possible (FIDO2 security key or passkey), as CISA recommends, using a short-lived Temporary Access Pass or equivalent issued only after verification.
  3. Restore access in stages: core productivity first, privileged roles only through just-in-time activation after a waiting period of [24] hours of clean monitoring.
  4. Keep enhanced monitoring on the account and similar roles for [30] days.
  5. Return the desk to normal mode only when sibling hunting is complete and the decision is logged.

7. Communication and notification

  • Service desk staff: brief all agents within [hours] with the caller's script, numbers and target names, and remind them they will be supported for refusing a request. See the Vishing and Help Desk Social Engineering Awareness Guide.
  • Affected users and managers: explain what happened and how to recognise follow-up contact.
  • Outsourced providers: require their incident report and confirmation of agent actions, per contract.
  • Regulators and law enforcement: legal assesses personal data duties under [GDPR Article 33 / sector rules]. CISA asks US victims of related ransomware to report to the FBI, IC3 or CISA; UK organisations report to the NCSC and Action Fraud as appropriate.
  • Assume monitoring: where the attacker may have reached email or chat, move incident communication out of band, as Co-op reportedly warned its staff.

8. Post-incident review

  • Replay the call with the agent and their manager, focusing on which control failed, not on the person.
  • Check verification design against NIST identity proofing guidance. See the NIST SP 800-63A-4 Identity Proofing Guide.
  • Measure time from call to report, report to revocation, and reset to attacker sign-in.
  • Learn from public cases: the MGM Resorts breach reportedly began with a single help desk call and ended with ransomware on ESXi hosts, and the Co-op cyber attack began with an impersonated password reset and moved to the Active Directory database.

Implementation guidance

  1. Map who can reset what across internal and outsourced desks, and remove privileged-user resets from front line roles.
  2. Build the detections in the "Where to look" table, starting with reset followed by new device sign-in within [60] minutes.
  3. Pilot heightened verification mode with one desk team, measure call handling time, then make it switchable for all teams.
  4. Rehearse quarterly with a red team call against a consenting test identity.
  5. Review metrics monthly and roll back any verification step that pushes users to unsafe workarounds.
MetricWhat it shows
Median time from suspicious call to SOC ticketWhether agents feel able to report
Time from report to session revocationContainment speed
Percentage of privileged MFA resets with second approverControl adoption
Resets followed by risky sign-in within 24 hoursResidual attacker success
Privileged users on phishing-resistant MFAReduction in reset value to attackers

Framework mapping

FrameworkReference
NIST CSF 2.0PR.AA-02 (identity proofing), PR.AT-01, DE.CM-03, RS.MA-02, RS.MI-01, RS.AN-03, RS.CO-02
NIST SP 800-53 Rev. 5IA-12 Identity Proofing, IA-2, AT-2, IR-4, IR-6
MITRE ATT&CK v20T1566.004, T1078, T1078.004, T1556.006, T1098.005, T1136, T1199, T1219, T1621

Standards and references

Related NHI Mgmt Group resources: Account Recovery and Help Desk Security Guide · Scattered Spider Threat Actor Profile · Vishing and Smishing Guide · MFA Fatigue and Bypass Guide · Identity Incident Response Playbook

PDF for subscribers

Take this guide with you

Every section, table and checklist in one PDF, with links to the sources and related guides. Subscribe to our weekly NHI & AI Identity Journal and we’ll email you a download link.

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 27 October 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

PDF Downloads for Newsletter Subscribers

Get the guide

Enter your details and we’ll email you a download link.

The full guide is free to read online. The PDF is a thank you for subscribing.

Fill in all fields marked * and tick the PDF box to continue.