Microsoft Services Hit by Major Global Outage

Microsoft services worldwide suffered a major outage, disrupting access to popular platforms including Outlook, Microsoft 365 and Teams and leaving millions of users unable to use them normally.

The services are gradually returning to normal.

The disruption began on Thursday, with around 130,000 users reporting problems on Downdetector, according to a report published on Friday (November 13).

Alongside Word, Excel, PowerPoint, Outlook and Teams, other affected services included SharePoint, OneDrive and the AI assistant Copilot.

Although the number of complaints later fell to around 13,000, many users continued to report difficulties sending and receiving emails and accessing their inboxes.

A Microsoft spokesperson said there was no indication of a cyberattack. The company believes the outage was caused by a technical problem with its servers.

Microsoft said several servers in its infrastructure had unexpectedly become unavailable. This placed additional pressure on the remaining operational servers, leading to connection disruptions across various services.

The company said the servers that had been restarted were now stable. Efforts were underway to bring the remaining servers back online and maintain services through alternative network arrangements.

Users are expected to regain normal access gradually as service capacity increases, Microsoft said.

The outage also affected AI services operated by OpenAI and Anthropic.

OpenAI said problems with Microsoft's connectivity systems were disrupting the creation of new connections and the use of some tools. Anthropic reported similar issues.

Mark Gregory, an associate professor in the School of Engineering at Australia's RMIT University, said the disruption may have resulted from a fault in Microsoft's network infrastructure.

He explained that when some servers become unavailable, the remaining servers face additional pressure, which can eventually cause further failures and allow the problem to spread.

To prevent similar incidents in the future, network configuration changes should be properly tested before being implemented, he said. Organisations that depend on cloud services should also maintain alternative arrangements.