The major Microsoft outage has now ended. The company states that the disruption to Microsoft 365 has been resolved. The issue affected Outlook, Teams, and Copilot for over 24 hours.
In its final status update regarding the incident, the company stated, “We have completed the work to address the root cause of the incident and have restored services.” The disruption impacted users in at least 74 countries, forcing companies, governments, hospitals, and airlines to rely on phone lines, paper records, and—in some offices—fax machines to continue their operations. The problem began on November 12.
The issue surfaced on November 12, when users in the United States reported being unable to access their Exchange Online mailboxes. Microsoft first acknowledged the problem at 5:54 PM GMT (11:24 PM IST). Just 13 minutes later, the company warned that users in any region could potentially be affected. Why did this happen?
The company has expressed deep regret for the disruption, clarifying that the incident was not caused by a cyberattack or AI-related issues. Microsoft attributed the Outlook outage to a previously implemented network upgrade.
According to a Microsoft spokesperson, the incident stemmed from an issue encountered during efforts to increase network capacity—a problem that emerged following recent network changes.
The company described it as a technical issue, unrelated to cyber threats or AI/agents. This clarification followed an intensive, overnight effort to resolve the situation. In earlier updates posted on X (formerly Twitter), Microsoft had initially cited “excessive load” in several regions as the cause. A few hours later, the company stated that there had been no increase in traffic.
Several servers suddenly stopped functioning after failing automated health checks. This shifted their workload onto the servers that were still operational, thereby increasing the pressure on them as well. Engineers eventually identified a part of the problem, which was linked to a core process on the affected servers—a process responsible for handling secure connections. At one point, Microsoft acknowledged that the measures they had taken did not provide the expected relief.
Why did it take nearly an entire day for Outlook and Teams to be restored?
The most challenging task was keeping the servers running once they were brought back online. Microsoft explained that as soon as traffic hit the servers that had been restored to service, they risked failing again. The company’s engineers worked on multiple fronts simultaneously.
They limited the types of traffic reaching the affected infrastructure, reverted some servers to their previous configurations, applied patches to authentication and routing components, and added virtual machines to increase capacity.
The situation was further complicated by email traffic. Mail servers outside of Microsoft’s network kept attempting to resend messages that had failed to deliver; consequently, as soon as the servers came back online, they were flooded with a backlog of queued emails.
To ensure that no messages sent during the outage were lost, Microsoft extended its email timeout window to 48 hours. The company stated, “Given the high level of demand, we restored services gradually and verified stability at each stage. While this took longer, it helped ensure a stable restoration of services.” On November 13, the situation first stabilized in the Asia-Pacific region, followed by improvements in Europe and North America. By the afternoon of November 13, Microsoft reported that email queues had significantly decreased and were nearing normal levels.
Who faced the outage?
Airlines, hospitals, and courts across at least 74 countries were affected by the Microsoft outage. In January, Microsoft stated that it had over 450 million paid commercial subscriptions for its 365 services; the scale of this reach was evident in the impact felt across the affected sectors.
In India, the outage affected the email support services of IndiGo and Akasa Air. Air India rerouted its emails to ensure customer messages continued to be received, and its flights operated as scheduled. At a major Indian brokerage firm, a backup system proved useful for Outlook, but handling client calls via Teams became difficult.
In the UK, NHS staff had to resort to using paper records, pharmacies warned of delays in processing prescriptions, and a judge in Manchester sent jurors home.
In Colorado, a sheriff’s office had to ask individuals to appear in person to post bail, as the online payment process relied on Outlook email notifications.
Some companies in Japan reverted to using faxes. OpenAI and Anthropic also encountered issues with their Microsoft 365 connectors, while the electric truck manufacturer Workhorse Group postponed the release of its quarterly results until November 16.
Some aspects may remain unclear to users: emails held in a queue display the time they were delivered, rather than the time they were originally sent. Therefore, a message dated November 13 might actually have been written on the night of November 12.
