Skip to content
Trending
BreakingLaptops & PCs1 min read

Microsoft 365 outage: What broke, why Outlook and Copilot failed, and why the fix couldn’t just be ‘reboot’

A dormant Windows server bug — deployed in July, missed in testing, triggered by traffic shifts — collapsed Microsoft 365 for nearly 24 hours. Outlook went dark. Copilot vanished. Teams disconnected. And no, AI didn’t do it.

Get this deal

Paid links. As an Amazon Associate, Hitechreports earns from qualifying purchases. We also earn a commission from eBay and Temu on purchases made through our links. Prices change; the story gives the price as reported when it was written.

By Boot Sequence

The Computing Desk · (34 min ago)

Reported fromoutlet, credited1
Photo via Wccftech

Microsoft 365 outage lasted 23 hours 50 minutes — from 4:20 PM UTC November 12 to 4:10 PM UTC November 13, and took down Outlook, Teams, Copilot, SharePoint, and OneDrive for users worldwide. The root cause was a Windows server defect deployed in July 2026 to Microsoft’s own Microsoft 365 networking gateway infrastructure, not consumer or enterprise Windows installations.

It lay dormant for over three months before triggering cascading failure across more than 80% of those gateway servers. No AI involvement. No breach. Just a lock contention bug in an internal Windows process that spiked CPU and memory use until servers froze, and stayed frozen after restarts because the flaw reactivated under load.

Microsoft’s labs missed it: their test hardware had fewer CPU cores, so the race condition never surfaced. Recovery required custom code, not rollback. The fix restored the original connection-handling logic. Webmail returned globally by 2:17 PM UTC on November 13.

Microsoft confirmed full resolution via its official Status X account on November 14. A final Post-Incident Report is due November 20. The company has since added traffic-distribution safeguards, expanded monitoring, and increased capacity, and extended email-delivery timeouts during the incident to prevent message loss. All queued mail was processed afterward.

This was a Microsoft 365 outage, not a Windows outage, not an Azure outage, not a security incident. It was a single-point infrastructure failure, amplified by reduced capacity (14% already offline) and a routine Teams failover that pushed extra load onto the same fragile gateways. Outlook users in India, the US, and across Europe were hit hardest.

Copilot access failed where it depended on Exchange Online identity or data routing, which was most places. Teams calls dropped. SharePoint links broke. OneDrive sync stalled.

Microsoft says the bug does not affect any other Windows deployment. Its next step: shipping the patched gateway code to all remaining nodes, quietly, this time.

Get this deal

Paid links. As an Amazon Associate, Hitechreports earns from qualifying purchases. We also earn a commission from eBay and Temu on purchases made through our links. Prices change; the story gives the price as reported when it was written.

Reported from

How this story was made

Written by the Hitechreports desk from the reporting credited above, with facts attributed to their original publishers. We do not test devices ourselves; anything about performance, battery life or cameras comes from the outlets that did. Prices are as reported at the time of writing. Editorial policy · Report an error

What did you make of it?

Filed by The Computing Desk

Laptops, desktops, processors, graphics cards and the operating systems that run on them.

More from this desk →

Be the first to comment

Join the argument. No password, just your email or a passkey.

    Microsoft 365

    Find on Amazon