Microsoft 365 services were disrupted for hours on Thursday, Nov. 12, 2026, with users reporting problems accessing Outlook and other workplace apps. Microsoft attributed the strain to servers unexpectedly becoming unavailable—not to an overall increase in traffic—and was still working to restore affected servers at 6 p.m. ET.
1
2
17
What users experienced
Reports of Outlook problems climbed to nearly 130,000 on Downdetector shortly after 1 p.m. ET. Users described trouble loading inboxes and sending or receiving email. The number of reports later fell, reaching about 38,000 just before 5:30 p.m.; these are reports submitted to an outage tracker, not a verified count of affected users.
1
5
The disruption extended beyond email. Microsoft 365 and Teams also drew reports, while coverage identified Exchange Online and SharePoint Online among services potentially affected. Reports also described disruption to Copilot and Defender for Office 365. The available evidence does not establish that every service was unavailable to every customer.
3
7
16
28
Microsoft’s initial warning concerned intermittent Exchange Online mailbox connectivity in the Americas. A later advisory described possible Microsoft 365 connectivity impacts across regions.
20
What Microsoft said caused the disruption
Microsoft said a number of servers had unexpectedly become unavailable, shifting load onto the remaining infrastructure. Its investigation pointed to a problem involving automated server health checks; Microsoft was testing a configuration change intended to address servers failing those checks.
7
17
Microsoft also said it had not identified a recent change correlated with the start of the impact. That did not rule out every possible contributing factor: a later update said a small number of configuration updates were still under investigation.
3
29
Recovery was gradual
Around 4 p.m. ET, Microsoft said some affected servers had returned to service and that it was working to extend the fix. It also tested an alternative health-check method on part of the infrastructure and prepared an automated process to detect and restart unresponsive servers.
5
10
29
That progress did not immediately end the disruption. In a 5:09 p.m. update, Microsoft said its actions had not provided the expected relief and that an overwhelmed core process on affected servers was causing capacity loss. At 6 p.m. ET, it was still working to return servers to service, with only limited recovery reported.
2
30
Microsoft had not published a confirmed restoration estimate by early evening. A scheduled time for another progress update was not a promise that service would be restored by then.
19
What the outage reports do—and don’t—show
Falling Downdetector reports indicated fewer people were submitting reports at those times, but they did not confirm that all customers had recovered. Contemporary reporting still described service disruption in the evening, and Microsoft said it was continuing recovery work.
2
5
Some Microsoft Q&A posts also reported that the service-status page was loading slowly. Those posts do not establish a confirmed outage of the status page itself.
24