On Thursday, November 12, a Microsoft 365 outage cut millions of Americans off from Outlook email and dragged down Teams, Copilot and other services that depend on it. Customers were affected for about 24 hours. Microsoft has now explained why: a months-old bug in its own network servers, not a cyberattack. Attackers still moved quickly to exploit the confusion, sending fake outage notices that warned staff their “held” email would be deleted unless they clicked a link.
What went wrong
Microsoft’s preliminary post-incident report, dated November 15, blames a hidden defect in a Windows fix built only for Microsoft 365’s networking gateway servers. These are the “front door” that connections to Exchange Online pass through. The fix began rolling out in July and ran without trouble for more than three months. Microsoft’s lab tests missed the bug because they used less powerful hardware than its live servers, and the defect is worse on newer machines with more cores.
Three things lined up on November 12:
- Extra traffic: About 15 hours earlier, Teams automatically failed over onto the gateway that Exchange Online uses.
- Less capacity: Nearly 14% of gateway capacity was already offline because of an unrelated fault.
- Peak hours: The bug was triggered during the US peak traffic window.
Microsoft’s automatic recovery then made things worse. As servers stopped responding, their traffic was shifted onto healthy servers, which were overloaded in turn. At the low point, more than 80% of gateway servers were unresponsive, and regional safeguards failed to stop the problem spreading worldwide. Teams, OneDrive and SharePoint could switch to a second gateway and recovered sooner. Exchange Online cannot, so email went down for nearly everyone.
| When (US Eastern) | What happened |
|---|---|
| Fri, Nov 13, 11:10 AM | Customers no longer actively affected; backlog still clearing |
| Fri, Nov 13, 9:17 AM | Webmail availability back to normal levels worldwide |
| Fri, Nov 13, 2:20 AM | Microsoft starts rolling out a fix that reverts the faulty change |
| Thu, Nov 12, 2:00 PM | Microsoft pauses all deployments and configuration changes |
| Thu, Nov 12, ~11:05 AM | Monitoring flags connection failures; Exchange Online outage confirmed |
| Wed, Nov 11, 8:07 PM | Teams traffic fails over onto the gateway Exchange Online uses |
Times converted from the UTC timeline in Microsoft’s report (Office Watch). Microsoft extended its email timeout to 48 hours so delayed messages would not bounce, and says incoming mail was held rather than lost. Its final report, with the full list of repairs, is due November 20.
Who was affected
More than 430 million people use Microsoft 365, and at the peak about 130,000 people in the US reported Outlook problems on Downdetector (Quartz). Across the US, the outage reached well beyond office email:
- Congress: House staff told Politico the outage ended “a day of legislative work” (Wikipedia).
- State and local government: In New Mexico, state and local offices, hospitals, colleges and two utilities were affected. Pasadena, Fremont, Burbank and Santa Barbara warned residents that city email was down (Insurance Business).
- Courts and public safety: Courts from Los Angeles to Texas warned that communications could be delayed, and the Los Angeles County Medical Examiner’s Office lost email. Some law enforcement agencies reported trouble processing bail payments.
- Business and AI tools: OpenAI and Anthropic both reported problems with their Microsoft 365 connectors. Anderson Economic Group put lost productivity at about $7 billion.
Outside the US, reports came from 74 countries, including hospitals in Canada and the UK’s NHS.
Not a cyberattack
Microsoft’s report confirms that the outage came from its own software. Throughout the incident it said it was not aware of any malicious activity. Attackers exploited the outage anyway. On November 13, Microsoft Threat Intelligence found a phishing campaign aimed at organizations across several US industries.
The emails:
- claimed that messages had been held during the outage and would be deleted within 24 hours unless the recipient clicked a link to restore them
- had subject lines starting “Outage” followed by a six-character code, and sender names such as “Outage Notification” or “Service Health” behind an organization-specific prefix, so they looked like they came from the company’s own IT team
- quoted the real Exchange Online incident number, EX1550543, to look official
Expect more variations in the coming days: fake IT help-desk calls, offers of “outage refunds”, and payment-change requests hidden in the backlog of delayed mail.
There is a quieter risk too. When company email goes down, staff often switch to personal email or messaging apps to keep working. That moves sensitive data outside the controls your security team relies on.
Planning for the next one
This was Microsoft’s third major Microsoft 365 outage of 2026, after incidents in January and September. Microsoft’s report also shows that redundancy has limits: Exchange Online relied on a single gateway system, so when that gateway failed, email had nowhere to go. Organizations should:
- Keep an out-of-band communication plan, such as phone trees or SMS alerts, that does not depend on Microsoft 365.
- Decide in advance which tools staff may use when email is down, so they do not improvise with personal accounts.
- Map where single points of failure sit across email, identity and collaboration. Hybrid and on-premises Exchange can add resilience, but only if it is patched and secured as carefully as the cloud.
Fortify Your Server with Messageware Security
Data breaches have increased by 72%, servers are compromised in under 90 minutes. Ensure you have multiple layers of security software protecting your Windows Servers.
Server Threat Guard (STG) for All Windows Servers: Next-gen server protection, providing detection, alerting, and response (MDR) to zero-day and server penetration cyber-attacks. No need to research complicated deployments and no learning curve to install and manage.
EPG Guard for Exchange Servers: Real-time security. Stop AD account lockouts, eliminate password attacks, intelligent GEO blocking, and prevent Exchange Server vulnerability probing.
Don’t leave your critical infrastructure vulnerable, be proactive and stay ahead of evolving threats.