Threat reportPhishingTL-2026-6395
Payroll Pirates Phishing Campaign Targets Microsoft 365 Financial Workflows (AiTM Session Token Theft)
Payroll Pirates voicemail-themed phishing emails route victims through Google Meet, Google Ads and Amazon S3 redirects to an attacker-controlled AiTM proxy that relays Microsoft's real sign-in flow and captures session tokens. Operators refresh stolen sessions about every eight hours via residential proxies and quietly collect payroll, invoice and banking email from payroll, HR, finance and admin users.
Key facts
| Identifiers | TL-2026-6395 |
|---|---|
| Also known as | Payroll Pirate, Payroll Pirates campaign |
| Severity | HIGH |
| Affected products | Microsoft Microsoft 365 / Entra ID / Exchange Online (user accounts of payroll, HR, finance and admin staff) |
| Attribution | Storm-2657 (low confidence) |
| MITRE ATT&CK techniques | T1078.004 Cloud Accounts, T1087.004 Cloud Account, T1090.002 External Proxy, T1114.002 Remote Email Collection, T1204.001 Malicious Link, T1480 Execution Guardrails, T1528 Steal Application Access Token, T1539 Steal Web Session Cookie, T1550.004 Web Session Cookie, T1557 Adversary-in-the-Middle, T1564.008 Email Hiding Rules, T1566.002 Spearphishing Link and 1 more in the attack flow |
| Motivation | FINANCIAL |
| Target sectors | health, education, manufacturing, government administration, professional services, finance, human resources |
| Target regions | North America, united states of america, canada, Europe |
| Detection rules | 9 (1 shown free) |
| Sources | 8 cited sources |
Detection coverage
As of 2026-11-19, Threadlinqs Intelligence publishes 9 detection rule(s) for TL-2026-6395 across Splunk SPL, Microsoft KQL and Sigma, covering 23 indicator(s) of compromise. The whole corpus is readable without an account. Every threat comes with one complete detection in Splunk SPL, Microsoft KQL and Sigma, free for everyone; a free account adds daily TL Creds to unlock the rest. Paid plans add a larger daily allowance, correlation and the MCP server. Threadlinqs MCP server · View plans.
Free detection: Payroll Pirates Redirect Chain (Google Meet / Campaign Manager / S3 / AiTM Domains) — Exploitation
Detects users who touch the Payroll Pirates (TL-2026-6395) voicemail-lure redirect chain in web proxy logs: meet.google.com/linkredirect with dest= pointing at adservice.google.com.ph/ddm/clk, the attacker S3 bucket amzn-5646353653563view, or the known idp.* redirectors and mslogin./msonline./msauth./office.* AiTM proxy domains (Storm-2657 / Storm-2755 overlap). The chain precedes Microsoft 365 session token theft via an AiTM proxy.
- Languages: Splunk SPL, Microsoft KQL, Sigma
- ATT&CK: T1566.002, T1204.001, T1583.001
- Data sources: proxy:zscalernss-web, defender_for_office365:UrlClickEvents, web_proxy
- Severity: high
- Confidence: high
Splunk SPL
index=proxy sourcetype="zscalernss-web" earliest=-24h
| eval url_l=lower(url)
| eval hop_meet=if(like(url_l,"%meet.google.com/linkredirect%") AND like(url_l,"%adservice.google.com.ph%"),1,0)
| eval hop_ads=if(like(url_l,"%adservice.google.com.ph/ddm/clk/%"),1,0)
| eval hop_s3=if(like(url_l,"%amzn-5646353653563view%"),1,0)
| eval hop_aitm=if(match(url_l,"^(https?://)?(idp[.](keyreniao|korminel|kualabemo)[.]com|mslogin[.]milocaroline[.]com|msonline[.]logicalineonline[.]com|msauth[.]monlinelogicaline[.]com|office[.]ofrecie[.]com)([/:?]|$)"),1,0)
| where hop_meet=1 OR hop_ads=1 OR hop_s3=1 OR hop_aitm=1
| stats min(_time) as first_seen max(_time) as last_seen sum(hop_meet) as meet_hops sum(hop_ads) as ads_hops sum(hop_s3) as s3_hops sum(hop_aitm) as aitm_hops values(url) as urls by user src_ip
| where aitm_hops>0 OR s3_hops>0 OR meet_hops>0
| convert ctime(first_seen) ctime(last_seen)Microsoft KQL
let aitm_hosts = dynamic(['idp.keyreniao.com','idp.korminel.com','idp.kualabemo.com','mslogin.milocaroline.com','msonline.logicalineonline.com','msauth.monlinelogicaline.com','office.ofrecie.com']);
UrlClickEvents
| where Timestamp > ago(24h)
| where (Url has 'meet.google.com/linkredirect' and Url has 'adservice.google.com.ph')
or Url has 'amzn-5646353653563view'
or Url has_any (aitm_hosts)
| project Timestamp, AccountUpn, Url, UrlChain, ActionType, IsClickedThrough, NetworkMessageId, IPAddressSigma
title: Payroll Pirates Phishing Redirect Chain via Google Meet, S3 and AiTM Domains
id: 7c1f3a52-4b8e-4f6a-9d21-5e0a8b3c1d01
status: experimental
description: Detects web proxy requests in the Payroll Pirates redirect chain, a Google Meet linkredirect with an adservice.google.com.ph click tracker destination, the attacker S3 bucket object, or known idp.* redirector and AiTM proxy domains (TL-2026-6395).
references:
- https://arcticwolf.com/resources/blog/payroll-pirates-strange-new-tides-in-business-email-compromise/
- https://threats.wiz.io/all-incidents/payroll-pirates-phishing-campaign-targets-microsoft-365-financial-workflows
author: Threadlinqs Team
date: 2026-11-19
tags:
- attack.initial-access
- attack.t1566.002
- attack.execution
- attack.t1204.001
- attack.resource-development
- attack.t1583.001
logsource:
category: proxy
detection:
selection_meet:
cs-host: 'meet.google.com'
c-uri|contains|all:
- 'linkredirect'
- 'adservice.google.com.ph'
selection_s3:
cs-host|endswith: '.amazonaws.com'
c-uri|contains: 'amzn-5646353653563view'
selection_aitm_hosts:
cs-host:
- 'idp.keyreniao.com'
- 'idp.korminel.com'
- 'idp.kualabemo.com'
- 'mslogin.milocaroline.com'
- 'msonline.logicalineonline.com'
- 'msauth.monlinelogicaline.com'
- 'office.ofrecie.com'
condition: selection_meet or selection_s3 or selection_aitm_hosts
falsepositives:
- Security researchers or sandboxes detonating the campaign URLs
- Legitimate Google Meet link redirects (excluded by requiring the adservice.google.com.ph destination)
level: highKnown false positives:
- Security teams or sandboxes detonating the published campaign URLs
- Legitimate Google Campaign Manager clicks (only the combined meet.google.com/linkredirect + adservice.google.com.ph chain alerts)
Tuning notes: Adjust the sourcetype/field names (url, user, src_ip) to your proxy vendor or the CIM Web datamodel. Extend the domain list as Arctic Wolf and TL_OSINT_Scan publish new idp.* redirectors; AiTM domains are typically under 10 days old on Hostinger so pair with newly-registered-domain enrichment. Use the S3 bucket name as a high-fidelity pivot.
- Payroll Pirates Voicemail Lure Email Delivery and Click — Exploitation · unlock with TL Creds
- Payroll Pirates Session Reuse Across Rotating Residential Proxies (~8h Refresh) — Post-Exploitation · unlock with TL Creds
- Payroll Pirates MailItemsAccessed with Outlook App ID and Undocumented API ID — C2/Collection · unlock with TL Creds
- Payroll Pirates AiTM Proxy Fingerprinting Endpoint and Lookalike Microsoft Hosts — Exploitation · unlock with TL Creds
- Payroll Pirates Inbox Rule Moving Mail to Deleted Items and Marking Read — Post-Exploitation · unlock with TL Creds
- Payroll Pirates Sign-ins from AS27176 and SRA-Observed Infrastructure — C2/Proxy Infrastructure · unlock with TL Creds
- Payroll Pirates Anomalous Outlook Sign-in Clients and OfficeHome 90014 — Post-Exploitation · unlock with TL Creds
- Payroll Pirates Microsoft Graph axios User Enumeration (Payroll/HR/Finance) — Lateral Discovery · unlock with TL Creds
Indicators of compromise
As of 2026-11-19, Threadlinqs Intelligence tracks 23 indicators of compromise for TL-2026-6395.
Sample indicators (3 of 23)
- IP address:
216.247.226[.]32 - Domain:
idp[.]keyreniao[.]com - URL:
hxxps://s3[.]us-east-1[.]amazonaws[.]com/amzn-5646353653563view/url
Defanged, TLP:CLEAR. Full set of 23 indicators: unlock with TL Creds.
Technical analysis
Arctic Wolf Labs (2026-08-06) observed hundreds of organizations targeted by email in July 2026, with successful intrusions across healthcare, education, manufacturing, government and professional services in the US, Canada and Europe. Wiz classifies the activity as a campaign by the named actor Payroll Pirates (initial access: end-user compromise; impact: data exfiltration). Arctic Wolf reports significant technical and behavioral overlap with the cluster Microsoft tracks as Storm-2755 ("payroll pirate" attacks against Canadian employees, reported April 2026); SRA and The Hacker News describe the overlap as tactical and do not claim attribution.
Initial access is a voicemail-themed email impersonating an automated call notification service, with a Microsoft logo and fabricated caller ID, date, duration and reference number. Subject format: "[Organization] :ATTN: Review messages. Ref id: [random string]". The link passes through a six-hop chain: (1) meet.google.com/linkredirect?dest=..., (2) www.google.com/url?q=..., (3) a Campaign Manager click tracker at adservice.google.com.ph/ddm/clk/..., (4) an HTML object in an Amazon S3 bucket, (5) an idp.* redirector returning HTTP 302, and (6) the AiTM proxy on a Microsoft-themed subdomain (mslogin., msonline., msauth.) or a misspelled office.* domain. Recently used domains were often hosted on Hostinger and were under 10 days old. Arctic Wolf notes the Google Meet/Campaign Manager chain also appeared in public sandbox submissions delivering EvilTokens, Sneaky 2FA and FlowerStorm, suggesting shared delivery infrastructure rather than a common operator.
The proxy stack (OpenResty/1.31.1.1 front end, x-powered-by: Express on the proxy root, PHP/8.2.32 on the fingerprint endpoint) first fingerprints the browser through /st_58200519/class_identifier.php (OS, screen, language, time zone, hardware concurrency, WebDriver status, WebGL vendor/renderer). It then geolocates the victim via api.country.is and stores the result in an rcfh_country cookie valid for seven days, before redirecting to the proxied Microsoft OAuth endpoints (GetCredentialType, /common/login, /common/SAS/BeginAuth, /common/SAS/EndAuth). The real MFA flow is relayed, and the authorization code and ID token are captured at the proxy callback before the victim is sent to the genuine Office 365 site. The proxy sometimes omits the OAuth nonce parameter, producing Entra error 90014 on OfficeHome.
Anomalous sign-ins follow within minutes: initial sign-ins show implausible mobile Safari/Chrome on Windows 10; recurring Outlook sign-ins use Firefox 131.0, Firefox 151.0 or Python Requests instead of Edge. Session maintenance begins 11-24 hours after initial activity and recurs about every 8 hours through rotating residential proxies (mainly anyIP), reusing the same SessionID while IP, ASN and location change. Post-compromise, Microsoft Graph (axios user agent) enumerates users filtered on payroll, hr, finance and human resources terms, then mailboxes are collected for payroll, invoice, payment, banking and benefits content. A limited subset of intrusions created inbox rules that move messages to Deleted Items and mark them read, with those interactive sign-ins coming from hosting infrastructure (AS27176, Datawagon LLC). Most intrusions showed no MFA-method changes, device registration, credential changes or outbound phishing. SRA's June 2026 case data shows Graph reconnaissance from Canadian residential ISP space, initial authentication from US mobile carriers, and broad Graph scopes (Directory.Read.All, Files.ReadWrite.All, Group.ReadWrite.All, Chat.ReadWrite, User.ReadWrite). Microsoft's Storm-2755 reporting describes the end goal as diverting employee salaries by changing direct-deposit details in HR platforms such as Workday. Wiz severity is not provided in the source; HIGH is an analyst assessment.
MITRE ATT&CK techniques by tactic:
Defense Evasion
T1078.004 Cloud Accounts; T1480 Execution Guardrails; T1564.008 Email Hiding Rules
Discovery
Command and Control
Collection
T1114.002 Remote Email Collection
Execution
Credential Access
T1528 Steal Application Access Token; T1539 Steal Web Session Cookie; T1557 Adversary-in-the-Middle
lateral-movement
Initial Access
Resource Development
Remediation
Immediate actions
- Revoke all active sessions and refresh tokens for affected accounts
- Reset credentials and re-register MFA methods
- Remove malicious inbox rules that move messages to Deleted Items or mark them read
- Audit payroll/HR platforms (e.g. Workday, ADP) over the full dwell period for direct-deposit and payment-election changes
- Review MailItemsAccessed (Bind) events for App ID 5d661950-3475-41cd-a2c3-d671a3162bc1 paired with API ID c999ed3e-27ae-4cb3-b3a2-46b056af63d3 to scope accessed mail
- Hunt for recurring ~8-hour sign-ins sharing one SessionID across multiple ASNs/locations
- Block listed phishing domains and review enterprise app / OAuth consent grants added during the compromise window
Workarounds
- Alert on Outlook sign-ins from non-Edge browsers (Firefox 131.0/151.0, Python Requests)
- Alert on Entra error 90014 on OfficeHome combined with user-agent/OS mismatches
- Alert on New-InboxRule operations that delete or hide messages containing direct deposit, HR or bank keywords
Longer-term hardening
- Deploy phishing-resistant MFA (FIDO2, passkeys, Windows Hello for Business, certificate-based auth) enforced via Conditional Access authentication strength
- Require compliant or hybrid-joined devices in Conditional Access and review exclusions and legacy authentication paths
- Enable Continuous Access Evaluation and sign-in frequency limits
- Ingest non-interactive sign-in logs, Unified Audit Log and Microsoft Graph activity logs into the SIEM
- Connect HR SaaS to Defender for Cloud Apps and alert on payment-election changes and new MFA device enrollments
- Run phishing awareness training and simulation for payroll, HR, finance and admin staff
Affected and fixed versions
- Microsoft — Microsoft 365 / Entra ID / Exchange Online (user accounts of payroll, HR, finance and admin staff)
Vulnerable versions: Tenants without phishing-resistant MFA and token-binding controls
Timeline
- Microsoft reports Storm-2755 "payroll pirate" attacks against Canadian employees (SEO-poisoned fake Microsoft 365 pages, AiTM session hijacking, direct-deposit diversion); date per secondary sources, Help Net Security coverage dated 2026-04-10
- Security Risk Advisors publishes multi-client case data on AiTM session hijacking and Graph reconnaissance (failed OfficeHome 90014 sign-ins, Firefox 142/131 takeover, Graph user enumeration)
- During July 2026 Arctic Wolf observes hundreds of organizations targeted by voicemail-themed AiTM phishing emails (month-level date per source)
- Arctic Wolf Labs publishes 'Payroll Pirates: Strange New Tides in Business Email Compromise' with redirect chain, domains, user agents and detection guidance
- Wiz Threat Center publishes the Payroll Pirates campaign entry (status Finalized); The Hacker News covers the Arctic Wolf findings
- Wiz Threat Center entry last edited
Sources cited
- Payroll Pirates Phishing Campaign Targets Microsoft 365 Financial Workflows (Wiz Threat Center) Wiz · Research · 2026-08-07
- Payroll Pirates: Strange New Tides in Business Email Compromise (Arctic Wolf Labs) arcticwolf.com · Vendor report · 2026-08-06
- Payroll Pirate Campaign: AiTM Session Hijacking and Microsoft Graph Reconnaissance Across Multiple Client Environments (SRA) sra.io · Research · 2026-06-25
- Investigating Storm-2755: Payroll pirate attacks targeting Canadian employees (Microsoft Security Blog) Microsoft · Vendor report · 2026-04-09
- Payroll Pirates Abuse AiTM Phishing to Compromise Microsoft 365 (HivePro) hivepro.com · Research
- Storm-2755's Silent Payroll Heist Targeting Canada (HivePro) hivepro.com · Research
- Microsoft 365 AiTM Phishing Hijacks Payroll and Finance Email (The Hacker News) The Hacker News · News · 2026-08-07
- Poisoned Office 365 search results lead to stolen paychecks (Help Net Security) Help Net Security · News · 2026-04-10
Drafted by the Threadlinqs research pipeline from 8 sources; published ; last updated —.