Microsoft struggles to restore Outlook as global outage continues

Microsoft was still struggling to restore its workplace services on Friday, November 13, more than 12 hours after a global outage disrupted Outlook, Teams and Copilot. The company described its recovery efforts as having achieved only “marginal success” and had not provided a timetable for full restoration.

The problems began around 16:00 GMT on Thursday, when users reported losing access to Outlook. Within roughly an hour, Downdetector had recorded complaints from 74 countries and territories. Microsoft initially identified problems in the Americas before acknowledging that users anywhere could be affected. The disruption continued through the US working day and remained unresolved on Friday morning in Europe, with some restored servers failing again.

The total number of affected users remains unknown. By about 07:00 GMT on Friday, Downdetector had collected at least 43,000 Outlook problem reports from 72 countries. That count offers only a partial measure because it includes people who submitted reports to the site. Government systems in Singapore and Japan were disrupted, along with communications at the Supreme Court of the Philippines. Australian organisations affected included the country's spy agency and federal police.

Outlook suffered the heaviest disruption, with users unable to reach mailboxes through browsers, apps and mobile devices. Some messages that were delivered arrived late. Other services experienced partial interruptions, and problems extended to integrations used by outside companies, including OpenAI and Anthropic. Teams chats and calls largely continued to function, while desktop Outlook users could potentially still view previously downloaded messages.

Microsoft has not identified the root cause. It said several servers unexpectedly became unavailable and that increased load across regions was contributing to connection problems, although traffic had not risen. On Friday, the company was reversing a network change without confirming whether it had triggered the outage. Charles Lamanna, Microsoft's executive vice president of Copilot, Agents and Platform, said late Thursday that there was no indication of malicious activity or a cyberattack.

Australia's Home Affairs Minister Tony Burke said Microsoft had also told him it did not believe a cyberattack was responsible. Speaking in Sydney, he said the company believed the situation had similarities to the faulty update behind the CrowdStrike disruption. Technology analyst and journalist Rupert Goodwins told Al Jazeera that an incorrect internal update can cause servers to restart or lose communication with one another.

Attempts to redirect traffic away from affected servers produced limited recovery, and Microsoft said some changes were not working as expected. Walter Pasquarelli, an affiliate researcher at the University of Cambridge's Bennett School of Public Policy, explained that failed automated health checks could shift traffic onto other servers and overload them. Goodwins said recovering servers could then be overwhelmed by requests from other machines trying to reconnect and synchronise files.

Microsoft said early Friday that it was checking server performance across regions and verifying that patches worked after infrastructure reboots. Services may return gradually as capacity improves and applications reconnect. The company has faced other disruptions this year, including a nearly 10-hour Microsoft 365 outage across North America in January and widespread Outlook access problems in late August.