Microsoft says it has finally restored its Microsoft 365 services after an outage that affected Outlook, Teams, Copilot, and other apps for more than a day. The company posted on X on Saturday, November 14, to confirm that services had been restored for the remaining affected customers and regions. The outage lasted well over 24 hours and affected businesses and government agencies around the world.

As we reported yesterday, the problems started with Exchange Online at around 11:17 a.m. ET on November 12. A wider Microsoft 365 outage followed at 1:26 p.m. ET. Microsoft had been trying to get the affected servers back online since then, but some of its recovery attempts were unsuccessful.

What caused the outage?

Microsoft initially suspected unusually high traffic was behind the problems. However, further investigation found that several backend servers had unexpectedly become unavailable, leaving the remaining ones struggling to handle the traffic.

The company tried restarting some of those servers, but they became unresponsive again once traffic returned. Microsoft later traced the problem to an earlier network capacity upgrade that was exposed by a subsequent network change.

Engineers eventually deployed a patch targeting an authentication and routing component. Microsoft also gradually restored traffic and worked through delayed emails. The company says it found no evidence of malicious activity and has ruled out AI as the cause.

Who was affected by the outage?

Aside from Outlook and Exchange Online, the problems affected Teams calendars, meeting scheduling, Copilot, SharePoint, Intune, and other Microsoft services. Some features, however, continued to work throughout the disruption.

According to Reuters, Britain’s NHS had to rely on alternative systems, while Indian airlines including Air India, IndiGo, and Akasa Air reported issues with email support. Financial institutions were affected as well. One Taiwanese firm even resorted to using fax to confirm trades. Downdetector recorded more than 7.4 million outage reports worldwide, although the figure doesn’t represent individual affected users.