Microsoft 365 Outage Over: Outlook, Teams and Copilot Restored After Day-Long Failure
Microsoft has closed the book on the Microsoft 365 outage that knocked out Outlook, Teams and Copilot for millions of users, marking both incidents “service restored” late on Friday night and ruling out a cyberattack as the cause.
The company posted the final entries for the two incidents — Exchange Online incident EX1550543 and Microsoft 365 advisory MO1550821 — at 11:15 p.m. and 11:17 p.m. Eastern time on Friday, November 13. Microsoft said the failure was a technical problem tied to a network capacity upgrade, not an attack and not related to its AI tools.
What Microsoft says caused the outage
In its final incident entries, Microsoft said the outage “was a technical issue not related to cyber or AI / agents” and that it “was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change.”
The trouble began when a large number of Microsoft’s own servers failed automated health checks and became unexpectedly unavailable. That put abnormal load on the servers that were still running, and the strain spread across Microsoft’s cloud. The company first described the problem as affecting the Americas, then said Microsoft 365 connectivity could be affected in any region, with Exchange Online the primarily impacted service.
Microsoft apologized to customers in its updates, saying: “We are deeply sorry for the disruption this incident has caused you and your business.” The company said it has addressed the underlying issue and restored services gradually, validating stability at each stage.
What went down
The outage that began on Thursday, November 12, hit email hardest. Users could not connect to Exchange Online mailboxes through Outlook on the web, the Outlook desktop app, Exchange ActiveSync or the programming interfaces that other apps use to reach mail. Email delivery was delayed, and mail flow between on-premises servers and Exchange Online was affected.
The damage spread well beyond email. Teams users could not schedule or edit calendar meetings, Copilot could not answer questions about Outlook, calendar or files, and features across SharePoint, Microsoft Graph, Defender, Purview, Power BI, Planner, To Do, Intune and Power Apps were hit. Microsoft also said it was working through a backlog of email messages that piled up during the event, and reporting based on the company’s updates said the delayed-email queue had been extended to 48 hours.
Microsoft lists the incident’s end as 11:10 a.m. Eastern time on Friday, though its own updates that afternoon still described recovery work as continuing. The incidents were only formally marked resolved late Friday night, more than a day after the failure began. The Verge’s Tom Warren reported that his sources link the outage to configuration changes in Azure Virtual Network, Microsoft’s cloud networking service, though Microsoft has not confirmed that detail.
Why this matters
The outage struck at the heart of the workday for businesses, schools and government offices that run on Microsoft 365, and it arrived in the same 24 hours as the FAA’s nationwide ground stop on Friday, which halted departing flights across the United States for hours after a communications failure. Two major American technology failures in a single day put a spotlight on how much daily life now depends on a small number of cloud and communications systems.
Related Articles
For Microsoft, the incident is also a credibility test. The company competes with Google and Amazon for the cloud business of the world’s largest organizations, and a day-long failure of its flagship productivity suite — followed by a cause that traces back to its own upgrade work — is the kind of event customers remember when contracts come up for renewal. The full post-incident report, which Microsoft has promised, will be closely read for exactly how a capacity upgrade went wrong.
What happens next
Microsoft said it will publish a preliminary post-incident report within two business days and a final report within five business days. Until that report lands, the mechanism behind the failure is documented but the full root cause is not.
Users who are still missing emails should check their inboxes and spam folders over the weekend as the backlog clears, and administrators should watch Microsoft’s service health page for any follow-up notices. Anyone still seeing errors should report them through their organization’s IT support, since Microsoft says it is monitoring for an extended period to make sure no further fixes are needed.
Frequently Asked Questions
Is the Microsoft 365 outage over?
Yes. Microsoft marked both incidents — Exchange Online EX1550543 and Microsoft 365 advisory MO1550821 — as resolved, with the status “service restored,” in final entries posted late Friday night, November 13.
What caused the Microsoft 365 outage?
According to Microsoft, the outage was a technical issue introduced by a problem during an earlier network capacity upgrade that surfaced after a recent network change. The company said it was not a cyberattack and was not related to AI or AI agents. A full post-incident report is still to come.
Which services were affected by the outage?
Outlook and Exchange Online were hit hardest, along with Teams calendar and Copilot features, SharePoint, Microsoft Graph, Defender, Purview, Power BI, Planner, To Do, Intune, Power Apps and Windows 365. Services that do not depend on Exchange Online mostly kept working.
Will delayed emails arrive?
Microsoft said it was monitoring the backlog of email messages that built up during the event to make sure they all process successfully. Reporting based on the company’s updates said the delayed-email queue had been extended to 48 hours.
Microsoft’s outage is over, but the questions it raised — about how a routine upgrade can take down email for millions of people, and how quickly a cloud giant can recover — will not be answered until the company publishes its full report.
Sources:
- StatusStack (Microsoft Service Health Status recap)
- MacRumors
- RansomNews
- ZeroHour
