Companies$MSFT

Microsoft says hidden July bug caused 24-hour global Outlook outage

Microsoft says a latent software bug from July, combined with higher demand and lost server capacity, caused a 24-hour global Outlook and Microsoft 365 outage.

Key takeaways

  • Microsoft (MSFT) says a software bug introduced in July caused a worldwide outage that disrupted Outlook and Microsoft 365 for about 24 hours last week. More than 80% of servers were at one point unresponsive. Microsoft has promised a final report with more details.
MSFT logo

Key points

  • Microsoft says a latent software bug from July, combined with higher demand and lost server capacity, caused a 24-hour global Outlook and Microsoft 365 outage.
  • At its lowest point, more than 80% of servers were unresponsive, and a separate issue let the failure spread across the network.
  • MSFT closed down 1.10% and is up 0.06% after hours, with Microsoft promising a final report with more details.

Microsoft says a software bug introduced in July went undetected until last week, when it helped trigger a worldwide outage that disrupted Outlook, Exchange and other Microsoft 365 tools for about 24 hours. The company's initial report, issued on Monday, identified three factors that combined to cause the failure: the hidden flaw, increased demand and reduced server capacity.

The outage reached far beyond corporate inboxes. The Guardian reported that millions of people worldwide said they could not access email, and that jails, hospitals, schools and police departments were hit by the loss of access.

The disruption began Thursday with US users reporting interrupted Outlook access, and Microsoft said it had restored service by Saturday. Microsoft's executive Charles Lamanna said the trigger came on Nov. 12 during peak U.S. business hours, when Teams traffic moved to a system that also supports email. Nearly 14% of that system's server capacity was already offline because of a separate issue. When servers went silent, Microsoft's systems automatically rerouted requests to healthy machines, and those got swamped as well. At the lowest point, more than 80% of servers were unresponsive. A separate problem also stopped safeguards from containing the failure within individual network sections.

Recovery was slow. Servers returning to service were immediately swamped by waiting traffic and failed again. Engineers eventually applied a fix and restarted groups of servers before returning them to service gradually. Teams, OneDrive, SharePoint and Office apps moved off the faulty infrastructure and came back sooner. Exchange Online, which powers Microsoft's email, could not make the same switch. Because the bug had been present for months, the company's normal rollback tools could not quickly undo it.

Lamanna said the flaw came from a fix built specifically for Microsoft 365 gateway servers and does not apply to broader Windows deployments. The software change had run worldwide for more than three months without problems. Microsoft's labs lacked the blend of traffic and hardware that later set the flaw off, which is why its tests failed to catch it. Newer, more powerful machines were hit harder, but testing had leaned on weaker hardware.

Lamanna wrote on X that the company was "deeply sorry for the impact to all our customers." In an earlier post, he said "This was not a cyberattack or an issue caused by AI tooling." Rep. Rich McCormick, R-Ga., used the outage to call for better cloud risk assessment, pointing to his Common-Sense Cloud Security Act. He wrote on X, "The question isn't just how we prevent the next failure."

The disruption also hit Microsoft's Copilot AI assistant. Fox Business reported that airlines including IndiGo and Akasa Air saw problems with customer-support email systems, while OpenAI and Anthropic reported issues with their Microsoft 365 integrations.

MSFT shares closed down 1.10% today and are up 0.06% after hours as of 5:24 p.m. ET.

Sources: Bloomberg, Fox Business, The Guardian

Sources