Microsoft 365 Outage Finally Fixed: Company Explains What Triggered Global Disruption (Representative image) Microsoft has restored its Microsoft 365 services after an outage lasting more than 24 hours and affecting users in at least 74 countries. The disruption hit Outlook, Teams and Copilot, forcing businesses, hospitals, government offices and airlines to rely on alternative communication methods, including phone calls, paper records and fax machines. In its final status update, Microsoft said, “We have completed our efforts to address the underlying issue of the incident and restored service.” The company apologised to affected users, saying it was “deeply sorry” for the inconvenience. It also maintained that the disruption was neither a cyberattack nor an issue involving artificial intelligence. The outage began on November 12, when users in the Americas started reporting problems accessing their Exchange Online mailboxes. Microsoft acknowledged the incident at 5:54 pm GMT (11:24 pm IST). Within 13 minutes, it warned that the disruption could extend to users in other regions. Microsoft has traced the incident to a problem linked to an earlier network capacity upgrade, which surfaced after a subsequent network change. A company spokesperson said the issue “was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change.” The spokesperson also clarified that it was “a technical issue not related to cyber or AI/agents.” The explanation followed a series of changing assessments from Microsoft. Its initial updates on X cited “elevated load” across several regions. The company later clarified that an increase in traffic was not responsible for the disruption. Instead, multiple servers had unexpectedly become unavailable after failing automated health checks. Their workloads were redirected to servers that were still functioning, placing additional strain on the remaining infrastructure. Engineers then identified a problem with a core process used by the affected servers to handle secure connections. The process became overwhelmed, leaving the servers unable to process requests. The recovery effort initially failed to deliver the expected improvement. Microsoft acknowledged at one stage that its actions “have not provided the anticipated relief.” Getting the servers back online was only part of the challenge. Some systems stopped responding again when normal traffic was restored, forcing engineers to work on stabilising the infrastructure before services could fully recover. Microsoft restricted certain traffic reaching the affected systems, rolled back some servers to an earlier configuration, patched an authentication and routing component, and added virtual machines to increase capacity. There was another problem waiting for the restored servers: a massive backlog of undelivered emails.