On 19 October 2026, Microsoft 365 had three separate incidents affecting Microsoft Teams and Microsoft 365 apps. The longest, a Teams town hall join failure, lasted 17h 30m and traced to a code regression on Microsoft's side.
Started
07:56 UTC
Duration
Lasted 17h 30m
Source
IsDown
Next time
Microsoft 365 is degraded. Is your product working?
An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.
50 monitors for your own websites. No card.
Affected components: Microsoft 365 apps, Microsoft Teams
What happened?
The first and longest incident started at 07:56 UTC: some users creating town hall meetings in Microsoft Teams with registration disabled found attendees redirected to a registration page instead, or saw the Join option unavailable entirely. Microsoft's updates trace a full diagnostic arc. At 07:58 the company was still confirming the behaviour. By 09:55 it had identified a recent service change as the cause, specifically a regression that removed the attendee Join URL when registration was disabled. At 11:52 Microsoft confirmed the regression and began building a fix. By 13:57 the fix was built and tested internally, but deployment to the live environment took until 16:15 and beyond. A 22:16 update said most affected users had the fix but full deployment was running later than Microsoft had projected. The incident closed after 17h 30m. Separately, starting at 09:03, some users could not access or protect content with Azure Rights Management across Microsoft 365 services, lasting 1h 32m; Microsoft said it restarted unhealthy service components and saw early recovery. A third incident from 09:27 affected sending messages in Microsoft 365 Copilot chat across the web app, Windows, Teams, Outlook and desktop apps, lasting 1h 4m, with Microsoft reverting a suspected service change. One user report in the window, filed from the United States of America at 20:43, described slow response times and problems logging in, which does not map cleanly to any of the three filed incidents.
Learning
None of these three incidents were the kind of failure your own uptime checks against Teams or Microsoft 365 endpoints would reliably catch, because each one was scoped to a specific internal workflow rather than a basic reachability failure. The town hall issue only affected users joining meetings with registration disabled, the ARM issue only affected encrypt or decrypt operations on content, and the Copilot issue only affected message sending through particular clients. Your product could keep working while any of these silently degraded a subset of functionality that a simple ping or login check never exercises. Microsoft's own root cause findings show the failure sat in application logic and service infrastructure on their side, not in anything reachable from outside. That is exactly the kind of fault where your monitoring, built around your own endpoints, stays green while real functionality is broken upstream. Over the past 90 days Microsoft 365 has logged 58 incidents with a median duration of 7 hours 15 minutes, so this pattern recurs often enough to plan for. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.