Microsoft Blames Capacity Update for 12-Hour Outlook Outage

Microsoft says a failed network capacity update, not a cyberattack, knocked out Outlook, Teams and the wider Microsoft 365 suite for more than 12 hours, hitting users across 74 countries. The company apologized on Friday and said it had fixed the root cause, though recovery would be gradual.

The numbers tell the scale. Downdetector, the platform that tracks complaints about digital services, logged problem reports from users in 74 countries. The outage ran past the 12-hour mark before Microsoft registered meaningful progress restoring its infrastructure.

What caused the failure?

In a statement, Microsoft said the problem originated in a failure during a previous network capacity update, which surfaced after a recent change to its infrastructure. Several servers became unexpectedly unavailable, pushing load onto the machines still running and triggering connection problems across regions.

The company ruled out two other explanations. It found no sign of malicious activity or a cyberattack, and it said the failure was not caused by a spike in user traffic.

That chain effect also explains why early recovery attempts fell short: some servers went unstable again as soon as they received user traffic. Microsoft said it has now identified and corrected the core problem, but stressed that recovery had to happen gradually to keep the services stable.

"We sincerely apologize for the disruption caused to our customers and partners," the company said.

Takeaway: one bad infrastructure change cascaded into a global outage, and Microsoft still has a deeper technical investigation ahead.

Who did the outage hit?

The disruption went well beyond consumers locked out of their inboxes. The failure hit the full Microsoft 365 stack — Word, Excel, PowerPoint, SharePoint, OneDrive — plus the Teams videoconferencing service and the Copilot AI assistant.

The operational fallout spread across sectors:

• In the United Kingdom, health services, police forces and schools reported problems; some institutions fell back on phone calls and manual procedures to work around the unavailable systems.
• In finance, the outage disrupted company operations and delayed corporate earnings announcements. One Taiwanese financial institution resorted to fax machines to confirm trades.
• Airlines warned of trouble with email-based customer service.
• Integrations used by the AI firms OpenAI and Anthropic with Microsoft 365 were also affected.

Takeaway: when one vendor's productivity stack goes down, hospitals, banks and airlines feel it within hours.

What happens next?

Microsoft said improvements appeared first in North America and Europe, with signs of recovery in Asia-Pacific as well. Traffic is being rerouted gradually, and access to Outlook inboxes has started to come back. Some users may still face sporadic interruptions and delayed message delivery.

The company has promised to publish more detail on the causes of the outage and the steps it will take to prevent a repeat. For the marketers, agencies and enterprises that run their day-to-day operations on Microsoft 365, that post-mortem — and the credibility of the fixes in it — is the real test of Friday's apology.