REDMOND, Washington — Microsoft has apologised after a major global outage disrupted Outlook and other Microsoft 365 services last week, leaving businesses, government offices, schools and healthcare organisations struggling with email and online communication across two working days.

Laptop and smartphone showing email service outage alerts against a cloud network and data centre background
Illustrative AI-generated image representing the global email and Microsoft 365 service disruption. It is not an official Microsoft image.

In an update published on Monday, November 16, Microsoft executive Charles Lamanna outlined three factors that combined to trigger the failure: a previously undetected software defect, increased network traffic and reduced server capacity.

Lamanna, Microsoft’s Executive Vice President of Copilot, Agents and Platform, said the company was deeply sorry for the disruption and recognised the impact on customers’ normal business operations.

Three Problems Combined To Trigger The Outage

According to Microsoft’s preliminary post-incident explanation, the outage began on Thursday, November 12, when a latent software defect caused networking gateway servers to become overwhelmed and unresponsive under certain traffic conditions.

The defect was connected to a software fix developed specifically for Microsoft 365 networking gateway servers. The change began rolling out in July and had been deployed globally for more than three months without an observed incident.

Microsoft explained that the underlying technical problem involved lock contention inside an internal process. Under heavy load, competing operations could become stuck waiting on one another, consuming server resources until affected machines stopped responding.

The second factor was a surge in traffic through the affected network gateway. About 15 hours before the outage, Microsoft Teams traffic was moved from an alternative gateway to the primary gateway that also supports Exchange Online, the service behind Outlook email. Microsoft described the failover as part of normal operations, but said it increased the load and changed traffic patterns.

The third factor was reduced capacity. Nearly 14 per cent of the gateway’s server capacity was already unavailable because of a separate issue from the previous day. Microsoft said the capacity was expected to return through routine maintenance and health processes, but its absence became important when traffic increased.

Microsoft said none of these factors alone would necessarily have caused the incident. Together, however, they exposed the dormant software defect during a peak period of United States traffic, causing requests to accumulate until servers stopped responding.

Why Recovery Took So Long

The disruption was made worse by the recovery process itself. As engineers brought servers back online, the waiting requests immediately overwhelmed them, causing some machines to become unresponsive again. Microsoft described the resulting surge as a major complication in restoring service.

The company also acknowledged that its normal automated recovery processes were not designed for such a widespread state of server unresponsiveness. Because the defective code had been introduced months earlier, ordinary rollback mechanisms and feature controls could not quickly reverse it.

Microsoft eventually developed an update that returned the affected code path to a previously known working state. Engineers removed groups of servers from service, applied the correction, restarted them and gradually restored capacity while checking stability.

Technology publication GeekWire reported that Microsoft’s own automatic traffic rerouting contributed to the cascading failure, as traffic moved away from unresponsive servers and placed additional pressure on others that contained the same defect.

Outlook, Teams And Other Services Were Affected

The incident affected Outlook and Exchange Online, as well as other Microsoft 365 services including Teams, Copilot and related cloud applications. Users reported difficulties accessing mailboxes, sending and receiving messages, managing calendars and conducting normal online work.

Reuters reported that the outage lasted more than 24 hours, with delayed email continuing to arrive after service began recovering. The disruption affected organisations in multiple countries and exposed how heavily businesses and public institutions rely on Microsoft’s cloud infrastructure for daily communication.

Reports of disruption came from sectors including healthcare, government, education, aviation and financial services. In some cases, organisations had to rely on alternative communication methods while Microsoft worked to restore access.

Microsoft Says The Outage Was Not A Cyberattack

Microsoft has said there was no indication that the incident resulted from malicious activity or a cyberattack. The company also said the disruption was not caused by artificial intelligence or AI agents.

The explanation points instead to an infrastructure failure involving software behaviour, traffic distribution and unavailable server capacity. Microsoft has rolled back the defective code path and described additional steps intended to reduce the risk of a similar incident.

What Microsoft Says It Is Doing Next

Microsoft says it has added server capacity and monitoring, and introduced tighter limits on how traffic is redistributed when parts of the network become unhealthy. The company is also reviewing its recovery procedures to improve its response to large-scale infrastructure failures.

A fuller post-incident review is expected on November 20, according to the preliminary explanation. That report is expected to provide further details about the failure and the measures Microsoft intends to take to strengthen the reliability of its services.

For customers, the incident is a reminder of the operational risks that can arise when essential communication systems depend on a shared cloud platform. Even services designed with backup gateways and automated recovery can experience widespread disruption when several weaknesses align at the same time.

Sources And Further Reading