Published: November 13, 2026
It started with the inbox
At 16:17 UTC on 12 November, Exchange Online stopped handing out mailboxes. Outlook on the web, the desktop app, phones and the REST API all missed. Microsoft's line on the service health dashboard was blunt: people cannot reach their mail, and delivery is delayed. Within two hours the company said any region could be caught in it. On Friday morning the incident was still open, past fifteen hours, and Microsoft had told customers the latest fixes were "not taking as expected."
Downdetector recorded more than 130,000 Outlook reports at the peak. Microsoft is tracking Exchange as EX1550543 and the wider Microsoft 365 estate as MO1550821. The live board is status.cloud.microsoft.
Microsoft has ruled out a traffic surge and said there is no sign of a cyberattack. A large set of servers became unexpectedly unavailable, and the work fell on whatever was still standing. A process that opens secure connections then overloaded, and those servers stopped taking requests. Machines that were brought back have fallen over again once real traffic hit them. The recovery is not holding.
What is actually broken
This is the list we are working from, set against Microsoft's own incident notes rather than a rumour thread.
- Outlook, on Office 365. Mailboxes are failing on the web, on the desktop and on mobile. Mail is queueing instead of arriving.
- Microsoft 365 admin center. Pages are failing to load. The control panel is a poor thing to lose in the middle of an outage.
- Microsoft 365 Copilot. Chat is not loading, and it cannot reach mail, calendars or files while Exchange is down.
- Microsoft Fabric. Labelled reports are failing on the way out to PDF or PowerPoint, on publish, and when a labelled pbix is opened in Power BI Desktop.
- Microsoft Graph. API calls that touch mail, calendar or contacts are timing out. Tools that manage Teams, channels or membership through Graph stall with them.
- Microsoft Intune. Device enrolment, check-in and compliance policies have been dropping in and out.
- Microsoft Planner. Plan and task operations are unreliable.
- Power Apps. Sign-in on the Power Apps and Field Service mobile apps is failing. The web apps are the way around it. Microsoft's advice for anyone already in the mobile app is to stay signed in, and not to clear the app data or reinstall until this passes.
- Power BI. Exporting a report that carries a sensitivity label to PDF is failing.
- Microsoft To Do. The app is not loading, and task operations are failing.
- SharePoint external sharing. Sending a file to someone outside the organisation is breaking, even where the tenant and the site both allow it. Files already stored in SharePoint still open if you go to them directly.
What kept running
Teams is split. Microsoft says ordinary chat, channel messages and joining a meeting are largely still up. Calendars, invites, meeting edits, and private or shared channels are not, because those sit on Exchange. Files already in SharePoint and OneDrive can still be opened. The Power Platform admin center has stayed up. This was not an Azure collapse.
That split is why a mail fault has emptied so many other desks. Planner, To Do, Copilot, Intune, Fabric and a long tail of outside tools all borrow Exchange or Graph. OpenAI has said new ChatGPT connections into Microsoft 365 are broken and the tools are not behaving. The same will be true of any integration that reads a mailbox or a calendar through Graph, including ones we have built.
How to reach us while this lasts
Our Microsoft mail is in the same hole. Do not wait on [email protected] today. Write to [email protected]. The notice at the top of this site says the same thing, and it stays until the mail is moving again.
We are watching the status board and will amend this post if the list of broken services changes. Microsoft has promised a post-incident report once the mitigation holds. It has not held yet.