Microsoft says its Microsoft 365 services are fully back online. The company confirmed the restoration early Saturday, November 14, ending a global outage that knocked out Outlook, Teams and Copilot for more than 24 hours and hit users in at least 74 countries.

“We have completed our efforts to address the underlying issue of the incident and restored service to the remaining impacted customers and regions,” Microsoft said in its final status update, posted overnight on its Microsoft 365 Status account.

The scale of the failure showed up fast in the numbers. Downdetector logged 138,058 user reports of Outlook problems in the United States at the peak, around 1 p.m. ET on Thursday, with Microsoft 365 itself drawing more than 47,000 reports. Those are user-submitted reports, not a count of affected customers, but they put this among the largest Microsoft outages in years.

What went down, and when

Trouble started on Thursday, November 12, when users in the Americas found they could not open their Exchange Online mailboxes. Microsoft’s service health record puts the Exchange incident start at 11:17 a.m. ET. The company first acknowledged the problem at 5:54 p.m. GMT, and 13 minutes later warned that users in any region could be affected.

From there it spread. Outlook stopped loading across web, desktop and mobile apps, and email delivery slowed to a crawl. Copilot Chat failed to load for many users, and Copilot lost access to work data, including emails, calendars and files. SharePoint Online, Microsoft Defender XDR, Intune, Planner and Microsoft Purview were also listed as affected.

One detail softened the blow a little. Core Teams messaging and online meetings largely kept working, according to Business Standard, so plenty of offices could still talk to each other while their inboxes sat dead.

Recovery came in stages. Asia Pacific was the first region to stabilize on Friday, November 13, followed by Europe and then North America. By Friday afternoon GMT, Microsoft said email queues were “significantly reduced and approaching normal levels.” The company also stretched its email timeout window to 48 hours to keep delayed messages from bouncing while the backlog cleared.

A network upgrade, not an attack

With a failure this big, the first question was whether someone had hit Microsoft. The answer, from the company, is no.

“This was a technical issue not related to cyber or AI / agents,” Microsoft said on Friday. “The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change.”

Charles Lamanna, Microsoft’s executive vice president for Copilot, Agents and Platform, confirmed there was no indication of malicious or cyber activity. He also said technical challenges made the recovery “more difficult than normal,” which helps explain why a fix that looked close on Friday morning kept slipping.

Microsoft apologized on Friday, before the final restoration. “We are deeply sorry for the disruption this incident has caused our customers and partners. We know the trust they place in our services and take that responsibility extremely seriously,” the company said in a post on X.

A full post-incident report is coming once Microsoft’s internal review is done. Administrators can follow the incident under references MO1550821 for the wider Microsoft 365 disruption and EX1550543 for the Exchange Online mailbox problem.

The outage reached beyond Microsoft

This one did not stay inside Microsoft’s walls.

OpenAI and Anthropic both reported problems with their Microsoft 365 connectors during the outage. That meant ChatGPT and Claude could not reach the work email, calendars and files they are wired into for paying customers. The AI assistants were fine. The pipe they depend on was not.

That dependency is the quiet subplot here. ChatGPT Wallet launched the same week, and Claude’s workplace integrations keep expanding. Each new connection makes the assistant more useful — and each one adds another service that goes dark when the underlying platform does.

The real-world damage went further than missed emails. Governments, hospitals, courts and airlines reported operational problems during the outage, and some offices fell back on phone lines, paper records and, in a few cases, fax machines, according to the Times of India.

Why this matters

A day without email sounds survivable until it happens to a hospital, a court or an airline. Microsoft 365 is the default workplace stack for a huge share of American business, and this outage showed how much of the working day now runs through one company’s servers.

It also showed the limits of the usual reassurances. Microsoft paused deployments, shifted traffic away from the affected infrastructure and restarted servers. None of it worked fast. The company said at one point that recovery was “not taking as expected,” and for hours on Friday the picture kept changing.

For IT teams, the lesson is an old one, relearned at scale: have a way to reach people that does not depend on the same system that just failed. For everyone else, the lesson is simpler. The cloud is not a place. It is a set of buildings, run by people, and sometimes a network change goes wrong in a way that takes a day to unwind.

Microsoft has not said whether affected customers will get service credits. The post-incident report, when it lands, should answer that — and say what changes so the next network upgrade does not take Outlook down with it.

Frequently asked questions

What caused the Microsoft 365 outage?

Microsoft says the outage was a technical issue introduced during an earlier network capacity upgrade, which surfaced after a recent network change. The company says it was not related to a cyberattack or to AI agents.

How long was Microsoft 365 down?

The outage began around midday on Thursday, November 12, 2026, and lasted more than 24 hours. Microsoft confirmed full restoration early on Saturday, November 14, with services coming back in stages across regions.

Was the Microsoft 365 outage a cyberattack?

No. Microsoft and its executive vice president Charles Lamanna both said there was no indication of malicious or cyber activity. The company traced the failure to its own network infrastructure.

Did the outage affect ChatGPT and Claude?

Yes. OpenAI and Anthropic both reported problems with their Microsoft 365 connectors during the outage, so ChatGPT and Claude could not reach users’ work email, calendars and files until Microsoft’s services recovered.

Sources: Reuters, BBC, Business Standard, Times of India, Financial Express, Microsoft.