Microsoft 365 outage cuts off Outlook, Teams and Copilot for hours

Reports started shortly before 1 p.m. ET on Thursday, November 12, 2026. People around the world could not open Outlook, on the web or in the apps.

Microsoft's own incident log puts the start of the main problem at 1:26 p.m. ET. It tracks the problem under two codes, EX1550543 for Exchange Online and MO1550821 for the wider Microsoft 365 trouble.

Exchange Online is the service that stores and delivers email for Outlook. When it fails, your inbox is like a post office with the doors locked.

The trouble did not stop at email. Microsoft said customers might be unable to connect to several services at once:

• Outlook: web and app access to mailboxes.
• Teams: chat and meeting trouble, though less than email.
• Copilot: Microsoft's AI assistant inside 365.
• SharePoint Online and OneDrive: the file and document services.
• Defender XDR: Microsoft's security tools.

Other services, including Planner, To Do and Intune, were also listed as possibly affected. The security feature Safe Links, which checks web addresses in Outlook and Teams, may have been unable to verify links.

Downdetector, a site that counts problem reports from users, showed more than 130,000 reports at the peak. By Thursday evening that had fallen to around 40,000 or fewer, but access was still not back for most people.

Microsoft also could not always reach its own status page. At one point its Microsoft 365 admin center, where businesses read outage notes, was not loading either.

Big organizations felt it. New York City's emergency management office told residents that anything sent by email to a city agency, or any online city service that relies on Microsoft, might be delayed. It pointed people to 311 for non-emergency needs and 911 for emergencies.

In real life If your office runs on Microsoft 365, Thursday afternoon probably meant no email, no way to open shared files and a lot of people asking each other on their phones whether it was just them.

Microsoft's explanation changed as it learned more. That is normal in a live outage, but it helps to see the steps in order.

• First: a portion of infrastructure was under "elevated load" across multiple regions, which was likely adding to the connection errors.
• Then: the load did not come from extra traffic. A number of servers had become unexpectedly unavailable, so the servers still running had to carry too much.
• Next: servers were failing automated health checks, the small tests that tell the system a server is fit to take work. A server that fails the test is taken out of service.
• Later: a core process on the affected servers was becoming overwhelmed, which caused a loss of capacity.

Capacity here means how much work the servers can handle at once. Lose enough servers and the rest cannot keep up, so more of them fail. That kind of chain reaction is a likely reason the outage grew.

The catch: Microsoft said at around 20:19 UTC that it had not found any recent change that lined up with the start of the problem. It said the failed health checks might come from a specific component that handles server health. A final root cause had not been published.

Microsoft's communications director, Ben Hope, said in a statement that the company was aware of the issue and was "primarily" seeing problems with Exchange Online. Microsoft also said it does not believe malicious activity caused the outage.

The company described a long list of steps over the afternoon and night. Early on it paused all deployments, so no new updates would be pushed out while it worked. It moved traffic away from the affected infrastructure and restarted servers to put them back to work.

Those first steps did not give the relief Microsoft expected. It said it was testing a configuration change to stop servers from failing their health checks, and it was weighing whether to undo a small number of recent configuration updates.

Around 6 p.m. ET, there was good news. Microsoft said it had begun returning a portion of the affected servers to service and that they were staying stable. It said it had also shifted traffic to other pathways and seen limited recovery in some regions.

It also streamlined its recovery process to bring the remaining servers back faster. In one location it put protective traffic limits in place to reduce the load on the affected servers, and it said it would expand them if they helped.

What's next: the good news did not last everywhere. Microsoft later said the restored servers can become unresponsive again once they receive traffic. It was working on a fix and preparing an automated process that keeps looking for unresponsive servers and restarts them.

Shortly after 11:30 p.m. ET, it reported partial recovery in the regions where it had made changes. It added that these actions were not giving sustained recovery, so it was reviewing what else it could do. Microsoft still had no time for full restoration, and the outage ran into Friday morning.

Earlier on Thursday there was a separate Teams incident, tracked as TM1550488. It affected people in North and South America with delayed or failed messages and trouble loading chats.

Microsoft blamed resource contention in its infrastructure, which means too many tasks fighting for the same computing power. It said that incident was fully fixed. It has not said whether the two problems are related.

It might seem odd that servers coming back does not mean everything is fine at once. The reason is the backlog.

While Outlook was down, mail kept arriving and apps kept trying to reconnect. When service returns, all of that piles onto the servers together. Microsoft warned that users may see a gradual recovery as their apps reconnect.

Your inbox may fill in chunks, or show nothing new for a while, even after you can log in. A reconnecting rush can also push fragile servers over again, which fits what Microsoft said about servers becoming unresponsive under traffic.

Why it matters: many companies keep email, chat, files and now AI tools in one suite. One shared fault can take them all away at the same moment, and there is no second inbox to switch to.

Outages are a gift to scammers. When people are anxious to get back into their mail, fake messages that look like official notices work better.

Expect texts and emails saying you must reset a password or click a link to restore your account. Microsoft will not ask you to do that because of an outage. If a message says it will, do not click it.

• Go to the source: type the address of the service yourself, or use the official app.
• Skip the links: do not use links in unexpected messages about the outage.
• Ask your IT team: if you use Microsoft 365 at work, check with them first.

Microsoft's status page and the admin center are the places to look. For businesses, the incident codes EX1550543 and MO1550821 will show the latest notes there. If you want a similar story on how Microsoft's subscription changes affect you, see our look at the new shared storage plans.

• Do not panic about lost mail: Microsoft has not said any was lost, but it expects a slow, uneven return.
• Use another channel for urgent things: call, text or use another chat app if you need to reach someone now.
• Be wary of reset messages: an outage is a favorite cover story for phishing, which means fake messages that trick you into giving up your password.
• Check the status page before you troubleshoot: if Microsoft says the fault is on its side, resetting your phone will not help.