Microsoft Details Storm-2755 AiTM Campaign Targeting Canadian Payroll

· Original article ↗

Summary

Microsoft details Storm-2755’s Canada-focused campaign using SEO poisoning and malvertising to steal Microsoft 365 credentials and tokens, hijack sessions, and target payroll. Microsoft reports one payment was redirected through Workday.

Key points

  • SEO poisoning and malvertising directed Canadian users searching for Microsoft 365 to a fake sign-in page at bluegraintours[.]com.
  • Storm-2755 used AiTM to capture session cookies and OAuth tokens. Microsoft observed a shift to axios/1.7.9 with the same session ID, following Entra error 50199.
  • Non-interactive OfficeHome sign-ins occurred about every 30 minutes as stolen sessions were maintained; some accounts also saw password and MFA changes.
  • Compromised accounts were searched for payroll and HR information. Emails about direct deposits targeted HR or finance staff, while inbox rules hid messages containing terms such as “bank.”
  • Microsoft reports one case in which the actor used a stolen session to change banking details in Workday, redirecting a payroll payment.
  • Microsoft recommends revoking sessions and tokens, removing malicious inbox rules, using phishing-resistant MFA and Conditional Access, and monitoring for unusual sign-ins.
  • The article includes hunting queries for suspicious inbox rules and Workday payment-account changes.

Article Details

Attack Vectors
  • SEO poisoning and malvertising positioned an actor-controlled domain in search results for Microsoft 365 queries, directing Canadian users to a fake sign-in page.
  • An adversary-in-the-middle attack proxied authentication and captured credentials, session cookies, and OAuth tokens; replayed tokens enabled access without repeated authentication and bypassed non-phishing-resistant MFA.
  • Compromised accounts were searched for payroll and HR information, and social-engineering emails with the subject “Question about direct deposit” targeted HR or finance staff.
  • Inbox rules moved messages containing terms such as “direct deposit” or “bank” to hidden folders, suppressing alerts from HR.
  • The actor used stolen sessions to access Workday and change banking details, redirecting a payroll payment in one reported case.
Defensive Notes
  • Immediately revoke compromised tokens and sessions, remove malicious inbox rules, and reset affected credentials and MFA methods.
  • Implement phishing-resistant MFA, Conditional Access policies, and controls that block legacy authentication; use session-lifetime policies to require reauthentication.
  • Use continuous access evaluation to help revoke access when risk conditions change, credentials are reset, or policies are enforced.
  • Monitor for anomalous user agents such as Axios, suspicious sign-ins and token replay patterns, and unusual inbox-rule creation; audit OAuth applications.
  • Use Microsoft Defender detections and SIEM monitoring to investigate anomalous activity, and alert on Workday payment-election changes.

Indicators of compromise

TypeIndicatorContext
URLhxxp[:]//bluegraintours[.]comActor-controlled malicious landing page used in SEO poisoning and malvertising to steal credentials and tokens.

MITRE ATT&CK

CVE

Threat Actors

Vendors

Products

Countries

Related Articles