Thousands of users across the world faced disruptions to Microsoft 365, Outlook, Teams and several other Microsoft applications for over 10 hours on Friday.

The outage also affected OpenAI and Anthropic's AI services through their Microsoft 365 connectors.

Reports of Outlook outages peaked at around 1.30 am on Friday, with approximately 1,38,058 reports. The number later declined to around 34,490 by 4.30 am.

In a statement issued at 3.30 am, Microsoft said it was working to restore the affected servers, adding that services were gradually recovering as users reconnected.

The company's investigation identified a process responsible for establishing secure connections on the affected servers as a key factor behind the disruption. The process was becoming overwhelmed, causing servers to stop processing requests.

In its latest update, Microsoft said it was preparing a configuration change to resolve the issue. The change was being tested on a limited number of servers before a wider rollout. The company was also restarting affected infrastructure and reversing a recent network configuration change that coincided with the beginning of the outage.

Microsoft introduced temporary traffic restrictions in certain locations and monitored server stability under reduced connection loads before gradually restoring additional traffic.

The company clarified that the disruption was not caused by an overall increase in traffic but by a reduction in the number of servers available to handle requests.

"We are aware and investigating reports of users experiencing issues with Microsoft 365 services, primarily issues with Exchange Online," a Microsoft spokesperson said.

Multiple Microsoft services affected

The outage affected several services dependent on Exchange Online and Microsoft's email infrastructure.

Users faced difficulties accessing their mailboxes through Outlook on the web, desktop applications and mobile devices. Email delivery was also delayed.

Microsoft 365 users reported problems accessing recently used files, while some pages of the Microsoft 365 admin centre failed to load.

Microsoft Copilot was also affected, with users experiencing difficulties loading Copilot Chat and receiving responses through web and desktop applications. Its access to work-related information, including emails, calendars and files, was also disrupted

Microsoft Teams continued to support core messaging and online meetings largely normally. However, features dependent on Exchange Online, including calendars, meeting scheduling and certain organisational chat functions, were affected.

Other impacted services included Microsoft Planner, Microsoft To Do, Microsoft Defender, Microsoft Purview, Microsoft Graph, Microsoft Fabric, Power BI, SharePoint Online, Universal Print and Windows 365.

Microsoft 365 Education features, including Assignments, Reading Progress and Speaker Progress, were also affected. Some Microsoft Intune services were listed among the disrupted products.

The company said features that did not depend on Exchange Online continued to function.

OpenAI, Anthropic services also hit

The disruption extended beyond Microsoft's own applications, affecting third-party platforms that rely on its infrastructure.

OpenAI said it was investigating an issue with Microsoft connectors that could prevent users from establishing new connections and cause certain tools to malfunction.

Anthropic also reported an issue with an upstream provider affecting its Microsoft 365 connector.

The outage had business consequences as well. Electric truck manufacturer Workhorse Group postponed its third-quarter results announcement from November 12 to November 16, citing business disruptions caused by the global Microsoft outage.

The incident comes weeks after Microsoft reported a disruption to its Azure cloud platform, which affected its productivity software suite and businesses across several industries worldwide.