November 13, 2026
Threat actors are exploiting the ongoing Microsoft 365 service disruption in a phishing campaign designed to steal Microsoft 365 credentials. The campaign began appearing in inboxes early on November 13, shortly after widespread Microsoft service interruptions affected Outlook, Exchange Online, Teams, SharePoint, Copilot, and other Microsoft 365 services.
The timing of the campaign is significant. By impersonating Microsoft and organizational support teams while users are actively experiencing legitimate service problems, the attackers create a highly believable pretext for requesting immediate action.
Our analysis has identified multiple variations of the campaign targeting both corporate and consumer email users. Messages contain links leading to credential-harvesting infrastructure designed to capture Microsoft 365 login information.
Campaign Overview
The phishing activity uses the Microsoft 365 outage as the central social-engineering theme.
One lure reviewed during our investigation claims that messages were held during the outage and will be permanently deleted unless the recipient takes action within 24 hours. The message instructs the user to select a prominent “Restore Messages” button.
Notably, the lure references EX1550543, an authentic Microsoft Exchange Online incident identifier associated with the November 12 disruption. Microsoft publicly used EX1550543 while investigating Exchange Online connectivity problems before tracking the broader service disruption under MO1550821.
Use of a legitimate incident identifier adds credibility to the message and demonstrates how quickly attackers can incorporate details from an unfolding event into phishing operations.
Figure 1 — Microsoft outage-themed lure observed November 13.
The message claims email was held following the Microsoft 365 disruption and directs the recipient to restore the messages.
From Outage Notification to Credential Theft
Selecting the action button directs the victim toward a malicious webpage designed to harvest Microsoft 365 credentials. Rather than relying on a single static phishing URL, the observed samples use rotating or unique URLs across multiple domains. Infrastructure observed during the campaign has included URLs associated with several different domains and services, including SendGrid and Newbooker-related infrastructure. This type of URL variation can make traditional reputation-based blocking more difficult because individual recipients may receive different links, domains may change rapidly, and malicious infrastructure can be replaced as it is discovered. The destination ultimately presents the victim with an Office 365 authentication experience intended to capture login credentials.
Multiple Sender Identities and Customized Impersonation
The campaign is not associated with a single sender domain or identity. Observed messages have originated from multiple email domains and use different sender identities. In some instances, attackers appear to customize the sender's display name or local-part based on the organization being targeted. For example, a recipient associated with ABC may receive a message appearing to originate from an identity such as: "ABC Outage Notifications". This lightweight personalization can substantially increase the perceived legitimacy of the message. Rather than impersonating only Microsoft, the attacker creates the appearance of an organization-specific support or outage notification. Another lure variation identified during the investigation uses a broader “DOMAIN Support-Desk” identity and warns that a security issue has been discovered on the recipient's account.
Figure 2 — Support desk-themed variation of the campaign.
The lure directs the user to “Take Action” following an alleged security review.
The similarities in delivery behavior, branding, infrastructure, and credential-theft objective suggest the outage-themed messages may be part of a broader support-desk impersonation campaign rather than a standalone attack.
Broad Recipient Targeting
Observed messages have targeted both enterprise and consumer email accounts.
In some cases, the same phishing message contained recipients from both corporate domains and consumer email providers. This suggests the campaign may be using broadly sourced recipient lists rather than focusing exclusively on users within a single organization or industry. At the same time, organization-specific sender customization demonstrates that the operators are capable of adding targeted elements when useful. The result is a campaign combining elements of both high-volume phishing and basic target-specific personalization.
Exploiting a Real-World Event
The campaign illustrates an important characteristic of modern phishing operations: attackers can adapt extremely quickly to events that create uncertainty or disruption. The Microsoft 365 outage began on November 12 and affected Exchange Online and several services that depend on it, including Outlook, Teams, SharePoint, and Copilot. Microsoft reported that some users could experience problems accessing mailboxes as well as delayed email delivery. Those conditions provide attackers with an unusually effective social-engineering opportunity. A message claiming that email was delayed, quarantined, or lost does not require the attacker to manufacture the underlying problem. Many recipients had already experienced it firsthand. The lure therefore takes advantage of three conditions simultaneously:
- Relevance: Users are aware that Microsoft 365 is experiencing problems.
- Urgency: Messages allegedly awaiting recovery will be deleted within 24 hours.
- Trust: Microsoft branding, authentic incident terminology, and customized support identities reinforce legitimacy.
Together, these characteristics reduce the likelihood that a recipient will pause to question the request.
Security Control Observations
Several samples reviewed during this campaign were successfully delivered to Microsoft 365 inboxes despite native Microsoft email security protections. The timing is noteworthy because Microsoft's own service reporting indicated that parts of its security ecosystem were also affected during the outage. Microsoft reported impact to Defender XDR and indicated that Microsoft Defender for Office 365 Safe Links could experience problems verifying URLs. At this stage, however, there is insufficient evidence to conclude that disruption to Microsoft's security services directly enabled delivery of these phishing messages. The campaign instead highlights the importance of maintaining layered email security controls, particularly during major cloud-service disruptions when both users and security infrastructure may be operating under unusual conditions.
Why This Campaign Is Effective
Outage-themed phishing is especially effective because attackers are exploiting something the victim already believes to be true. Users experiencing delayed email or Microsoft 365 connectivity problems may reasonably expect communications from Microsoft or their internal IT team. A message offering to recover mail therefore fits naturally into the user's current experience. The use of a legitimate Microsoft incident identifier further narrows the gap between the legitimate event and the fraudulent message. Rather than creating an entirely fictional scenario, the attackers simply provide a malicious solution to a real problem.
Recommended Actions
Organizations should alert users to outage-themed phishing attempts and emphasize that unexpected email-recovery or account-security notifications should not be trusted solely because they reference a legitimate service incident.
Security teams should also consider:
- Reviewing messages referencing EX1550543, Microsoft 365 outage notifications, message recovery, or similar themes.
- Investigating recently registered or low-reputation domains embedded in outage-related messages.
- Examining messages using organization-specific variations of terms such as Support Desk, Outage Notifications, Security Check, or Message Recovery.
- Monitoring authentication telemetry for unusual Microsoft 365 login attempts following interaction with suspicious messages.
- Resetting credentials and revoking active sessions for users confirmed to have entered credentials into phishing pages.
- Enforcing phishing-resistant MFA wherever possible.
- Using layered email detection capable of analyzing message content, sender behavior, URLs, landing pages, and post-delivery threats rather than relying exclusively on static URL reputation.
Indicators and Detection Opportunities
Because the campaign uses rapidly changing URLs and multiple sending domains, individual indicators may have limited longevity. Behavioral characteristics may provide more durable detection opportunities, including combinations of:
- Theme: Microsoft outage, delayed email, held messages, account security, support desk
- Call to action: Restore Messages, Take Action, review held mail
- Credential target: Microsoft 365 / Office 365 credentials
- Impersonation: Microsoft or organization-specific IT/support identities
- Infrastructure: Rotating or recipient-specific URLs across multiple domains
- Targeting: Corporate and consumer email recipients
- Technique: Credential phishing through embedded links
The activity is consistent with MITRE ATT&CK T1566.002 – Phishing: Spearphishing Link, where an adversary sends a message containing a link intended to direct the victim to attacker-controlled infrastructure.
Conclusion
Major service disruptions create more than an availability problem. They also create an opportunity for attackers. Within hours of the November 12 Microsoft 365 disruption, phishing messages were observed using the outage as a pretext to convince recipients to visit credential-harvesting pages. The campaign combines real incident information, Microsoft branding, urgency, customized sender identities, and rapidly changing URL infrastructure to make malicious messages appear credible. Organizations should expect attackers to continue exploiting high-profile technology outages and other rapidly developing events. For defenders, the lesson is equally straightforward: when a major service disruption occurs, increased phishing activity should be treated as a predictable secondary threat—not an unrelated event.