Microsoft is struggling to restore services following a global outage that disrupted Outlook, Teams, Copilot, and several other Microsoft 365 applications.
The company’s initial recovery measures have not delivered the anticipated improvements, forcing it to explore alternative ways to restore its infrastructure.
The disruption began on Thursday, November 12, with Downdetector receiving reports from 74 countries and territories.
In its latest update at 11:44 am IST, Microsoft said it was evaluating additional traffic controls, including throttling, protocol adjustments, and failover routing, to improve infrastructure stability. It was also expanding internal testing to improve traffic monitoring and identify failures.
Earlier, the company had begun scaling up virtual machines to increase capacity and reduce server loads. However, it encountered problems synchronising network changes because some machines remained stuck in an abnormal state.
Microsoft said it was not aware of any malicious activity causing the incident. The company plans to publish a post-incident report detailing the root cause and corrective measures after mitigation.
Downdetector recorded nearly 130,000 reports of Outlook-related problems in the US alone. Users reported difficulties accessing their inboxes, sending and receiving emails, and connecting to workplace applications.
Other affected services included Microsoft Word, Excel, PowerPoint, OneDrive, and SharePoint, although the extent of disruption varied across applications.
The disruption primarily originated in Exchange Online, Microsoft’s cloud-based email infrastructure that powers Outlook.
Microsoft initially attributed the connectivity problems to elevated loads across its infrastructure after a large number of servers became unexpectedly unavailable. The resulting loss of capacity shifted traffic onto the remaining servers, creating further strain.
The company subsequently identified an issue with a core process responsible for establishing secure connections on affected servers.
The process was becoming overwhelmed, causing servers to stop processing requests. Microsoft also identified a recent network configuration change that correlated with the start of the incident, although it has not published a confirmed root cause.
The outage extended to services dependent on Exchange Online. In Teams, calendar access, meeting scheduling, and certain organisational chat features were affected, while core messaging and online meetings remained largely functional.
Direct access to files stored in SharePoint remained available, although features dependent on Exchange Online were disrupted. Other affected services included Microsoft Defender XDR, Planner, To Do, Purview, Intune, Power BI, and Microsoft Graph.
The disruption also affected third-party AI services. OpenAI reported problems with its Microsoft 365 connectors, warning that new connections could fail and connected tools might not work as expected.
Anthropic separately reported failures involving mail, calendar, and search functions in its Microsoft 365 connector, along with disruptions to Claude for Outlook.
Microsoft initially attempted to redistribute traffic across its global infrastructure, restart affected servers, and redirect connections to alternative pathways. However, some servers became unresponsive again when traffic was restored.
The company subsequently restricted certain types of traffic in affected regions, rolled back network configuration changes, and began rebuilding some servers using an earlier software configuration.
