It started with a send button that did nothing. On Thursday afternoon, people across the world clicked send on an email and watched it sit there, unsent. Outlook would not load. Teams would not connect. For millions of workers, the tools they use every single day had simply stopped.
More than a day later, the silence is over. Microsoft said early on Saturday that its Microsoft 365 services are fully restored, ending one of the longest outages in the platform’s history. Email is flowing again, meetings are back, and the company’s AI assistant Copilot is online.
The outage touched more than 70 countries and territories, disrupted governments, hospitals, courts and airlines, and produced millions of failure reports. Now the questions shift from when it will end to why it happened at all.
For Microsoft, the end of the outage is only the start of the explanation. The company has apologised, and it has promised a full report. But the details it has already shared point to a failure that began with an upgrade meant to make the network stronger.
What happened on Thursday
The first signs appeared on Thursday, November 12, at around 4 p.m. GMT. Users reported that Outlook would not open, and within about an hour the problem had spread far beyond one app. Exchange Online, the cloud hub behind Microsoft’s email and calendars, was struggling.
Downdetector, the service that tracks outages from user reports, logged problems from 74 countries and territories, including Israel, according to the Jerusalem Post. Reports kept climbing through the afternoon, and by the evening it was clear this was not a brief glitch.
More than 130,000 users reported issues on Thursday alone. At its peak, Downdetector logged nearly 20,000 reports a minute, more than 1,500 times the normal level. In total, the tracker recorded over 7.4 million reports worldwide, according to The Straits Times.
Microsoft’s own status page told the same story. Exchange Online features were degraded, and with them went Outlook, Teams, Copilot and a long list of connected services, from SharePoint to OneDrive and the Microsoft Store.
Copilot, Microsoft’s AI assistant, went down with the rest. That detail mattered, because it showed the outage was not limited to email. The same failure touched the company’s newest product and its oldest ones at the same time.
By early afternoon, the problem had a name and a scale. This was not a glitch in one app. It was a Microsoft 365 outage that pulled down Outlook, Teams, SharePoint and Copilot, and it would take Microsoft the rest of the working day to start putting it back together.
The scale of the Microsoft 365 outage
Outlook and Teams are two of the most widely used workplace applications on the planet, with hundreds of millions of users. When both go down at once, the effect is not a minor inconvenience. It is a working day that cannot start.
Government agencies in Australia, Japan and the United States reported problems, as did healthcare services, courts and airlines. Businesses could not send email, manage calendars or host virtual meetings. For many, the outage swallowed a full working day and part of the next.
The disruption lasted more than 24 hours. That makes it one of the longest outages in Microsoft 365’s history, and unusually long for business email. Most cloud outages are measured in hours, not days.
The numbers only tell part of the story. Behind every report was a person who could not do their job: a doctor waiting for a referral, a lawyer missing a filing deadline, a traveller unable to check in for a flight. The outage was measured in reports, but it was felt in missed work.
What caused the Microsoft 365 outage
Microsoft has now said what went wrong, and the answer is not what many people first feared. The company said the incident was introduced by an issue during an earlier network capacity upgrade, which surfaced after a recent network change.
The upgrade left a large number of servers unexpectedly unavailable. As systems tried again and again to reach them, the retries piled up into a cascade that overwhelmed the infrastructure that was still working. In short, the network change broke the servers, and the retries finished the job.
Microsoft was quick to rule out the two explanations people reach for first. The company said the outage was a technical issue, not related to cyber activity or artificial intelligence agents. Charles Lamanna, Microsoft’s executive vice president for Copilot, Agents and Platform, wrote on X late on Thursday that there was no sign of malicious activity.
How Microsoft brought the services back
Recovery came in stages. On Friday, Microsoft said its consumer services had fully recovered at 4:10 p.m. UTC, while its Microsoft 365 enterprise services were still on the way back. The company restored services gradually, checking stability at each stage because of the scale of demand.
Early on Saturday, the company confirmed the job was done. “We have completed our efforts to address the underlying issue of the incident and restored service to the remaining impacted customers and regions,” Microsoft posted on X, through its Microsoft 365 Status account.
The service health page agreed. By Saturday morning, it showed that all products were operational, including Microsoft 365 for business and enterprise customers. Microsoft added that it was still monitoring the backlog of email messages that had built up during the outage.
The slow pace frustrated some users, but Microsoft said it had little choice. Restoring everything at once risked overloading the same systems that had just failed. Each stage had to be checked before the next one began, and the email backlog had to be worked through in order.
Anyone who sent an email during the blackout should check that it actually arrived. Messages queued during the outage are being delivered, but a day-long backlog takes time to clear, and some users may still see delays.
A familiar pattern for Microsoft
This was the second disruption to Microsoft 365 services this year, after an outage in January 2026. It also follows an outage of Microsoft’s Azure cloud platform in October last year, which hit the same productivity tools and industries worldwide.
The timing made it worse. The outage began just two days after Microsoft’s November Patch Tuesday, the monthly security update. Microsoft has not linked the two events, but the week has now delivered two separate headaches for the company’s customers.
And Microsoft was not the only company having a bad week. A zero-day crisis at remote-management vendor N-able saw 33 flaws published and production servers mined for cryptocurrency, a reminder that the whole software supply chain is under strain.
“The bigger worry is that this is not the first Microsoft 365 outage in 2026,” one analyst told The Straits Times. For a platform that businesses treat as essential infrastructure, even one day-long outage raises hard questions about resilience.
What happens next
Microsoft has promised a full post-incident report once its review is complete. That report, tracked under incident number MO1550821, should explain in detail how a routine network capacity upgrade took down email for a day.
For now, the immediate advice is simple. Check your sent folder for anything important from Thursday and Friday, confirm that meetings scheduled during the outage are still in place, and watch for the post-incident report if your organisation needs to explain the disruption to its own customers.
There is a wider question too. As more of the working world runs on a handful of cloud platforms, a single failure can now stop work in dozens of countries at once. The Microsoft 365 outage is a reminder that the cloud is powerful, but it is not invincible.
The outage is over, and the inboxes are full again. But for the millions of people who spent a day unable to send a single email, the memory of a quiet Thursday afternoon will last a lot longer than the outage itself.



