Microsoft Act now
Microsoft 365 outage that started Nov. 12 still lacks a full fix
A Microsoft 365 breakdown that began Thursday, Nov. 12, knocked out Exchange Online and services that depend on it, including Outlook, Teams, and Copilot. Microsoft said it saw no malicious activity and that the underlying fault was fixed with restoration under way, but on Friday it still had no deadline for a complete return.
Microsoft 365 suffered a global outage from Thursday, Nov. 12, after Exchange Online failed and dependent tools stopped for organizations. Into Friday, Microsoft still gave no full-restoration deadline. It said it knew of no malicious activity behind the incident and that the underlying issue was fixed, with restoration under way.
Microsoft's service-health dashboard timed the start at 16:17 UTC. The Register said about 18:00 UTC and later that the outage had run past 27 hours while fixes kept failing. Outlook on the web, desktop, and mobile could not open mailboxes over MAPI, Exchange ActiveSync, or REST, and delivery was delayed; Apple Mail failed for the same reason.
Affected services included Teams, Copilot, Copilot 365, Defender for Office 365, Defender XDR, Intune, and SharePoint Online, though products with no Exchange tie stayed up. Most Teams calling, meetings, and messaging still worked, with errors on scheduling and calendars. Copilot could miss work mail and files, while files stored directly in SharePoint stayed available.
Microsoft blamed servers failing automated health checks. It found no causal system change, but early checks pointed to a component that processes server health, while the status page weighed reverting a few recent configuration updates. An overwhelmed core process made servers stop responding and lose capacity. Near 22:00 UTC an alternate check restored a subset, without the expected relief. Isolation, restarts, traffic tests, and extra virtual machines gave only partial recovery that faded; some mitigations were not taking as expected. In MO1550821, Microsoft said many servers were unexpectedly unavailable, not that traffic overall had risen, and rebalancing pushed routing-service requests onto remaining back-end machines. On Friday it reported improving mailbox access in parts of Asia-Pacific, encouraging recovery in Europe and improvement in the Americas, and it raised the email timeout to 48 hours. Intermittent failures could continue.