The disruption had affected certain Microsoft 365 services worldwide, forcing some organisations to use fax machines while recovery progressed in phases
The global outage of certain Microsoft 365 services that began on 12 November and lasted for days, has been attributed by the firm to failing servers, not a cyberattack or staffing problems.
The outage had disrupted email, workplace collaboration and other Microsoft cloud services for businesses and governments worldwide. Microsoft said recovery was progressing in phases, with some services stabilizing but capacity not yet back to normal at that time.
The cause of the outage was announced around some factors:
- Several servers had unexpectedly become unavailable, placing excessive load on the remaining infrastructure.
- Preliminary investigations had found that failed automated health checks may have originated in a component that processes server health.
- The firm had not identified a recent change that could have triggered the failure, but had taken steps to prevent additional servers from being affected while restoring those already offline, according to one report.
- There was no indication of malicious or cyber activity, and the firm has also rejected suggestions that the incident was linked to its recent voluntary buyout program.
Lengthy disruption
The disruption lasted more than 12 hours and was among Microsoft’s longest service outages, according to Reuters. The fault had affected Exchange Online-dependent features, including email, while some services that did not rely on Exchange Online continued to operate.
Microsoft had initially applied fixes to affected servers, but later said those adjustments had not produced a sustained recovery. Outage tracker Downdetector had recorded reports across 74 countries and territories within an hour. In the United States, Outlook-related reports had peaked at nearly 140,000.
The outage had forced some organisations to revert to older communication methods. A Taiwanese financial company reportedly used fax machines to confirm trades, while at least one European bond repurchase was canceled. Also:
- European exchange operator Euronext reported email disruptions affecting its listing teams.
- British insurer Aviva had activated backup arrangements for its earnings calls
- Danish shipping firm Maersk had warned customers of possible delays to email communications and updates.
- US electric-truck maker Workhorse Group had had to reschedule its earnings announcement to 16 November 2026.
Recovery
On 13 Nov 2026 Microsoft announced that recovery actions were continuing, and that it was seeing incremental improvement across several workloads. It said service health remained variable and capacity had not fully returned to baseline.
Later, the firm later said services had fully recovered across Teams, Outlook and its other products. Australian minister Tony Burke said Microsoft had compared the incident with the 2024 CrowdStrike outage.