News · Published 2026-11-12
Microsoft 365 Outage 12 Nov 2026: What To Do
Exchange Online, Teams and Copilot went down on 12 November 2026. What Microsoft said, what to do, and monitoring that survives a mailbox outage. Not tested.
Update, 13 November 2026 (checked about 19:50 UTC): Microsoft now states a cause in its own words. Its status feed, newest post authored 17:57 UTC (12:57 pm ET), says: "The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions." It also says "This was a technical issue not related to cyber or AI / agents". The same post says services were restored gradually, "validating stability at each stage", and that Microsoft will "provide a full post-incident report once our review is complete". The feed status still reads "Service degradation" and the word "resolved" does not appear in it or in Microsoft's posts on X; the post promises a next update by 20:00 UTC. The 48-hour email timeout window is still in place as of Microsoft's 12:09 UTC (7:09 am ET) post on X. No post-incident report has been published yet. Earlier updates, including the 07:29 UTC patch Microsoft described as a mitigation, are in the timeline below.
TL;DR
- What broke: on Thursday 12 November 2026, Exchange Online mailbox access failed for many users, first in the Americas, then Microsoft said users in any region could be affected. Microsoft listed Exchange Online, Teams, SharePoint Online, Microsoft 365 Copilot and Microsoft Defender XDR. Its status feed (17:57 UTC post on 13 November) lists about twenty affected services, among them Microsoft Graph, Intune, Planner, Power Apps, Purview, Fabric, Windows 365 and Universal Print: see https://status.cloud.microsoft/.
- Microsoft's explanation so far: not a traffic spike. It said "a number of servers" were "unexpectedly unavailable", which pushed load onto the remaining infrastructure across several regions. It paused all deployments, moved traffic away from the affected systems and restarted servers.
- Microsoft says it was not an attack. Its status feed post authored 17:57 UTC on 13 November says "This was a technical issue not related to cyber or AI / agents", and that the incident "was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change". No threat actor has claimed it, according to the coverage we read.
- Recovery is gradual, and Microsoft has not said it is over. As of our check at about 19:50 UTC on 13 November the status feed still says "Service degradation", and neither the feed nor Microsoft's posts on X say resolved. Microsoft says it is restoring "remaining customers and regions". One outlet reported service back after about an hour; Microsoft's posts contradict that.
- What to do: do not change your tenant, your DNS or your mail routing. This is on Microsoft's side. Tell your users, watch the official channels and keep an independent check running, so you are not relying on Microsoft to tell you that Microsoft is down.
How this was checked
On 12 and 13 November 2026 we read Microsoft's own posts on X (@MSFT365Status), its status feed, an incident tracker (BrokenToday, secondary) and reports from BleepingComputer, Techaeris, Engadget, XDA Developers, ABC News and ynetnews. Microsoft's posts are the primary source for the table below. Where the sources disagree (the start time, whether it was over), this post shows both.
What happened
All times are US Eastern (ET, UTC minus 5) on 12 November unless marked. Microsoft's posts on X are the source unless another is named.
| Time (ET) | What was reported | Source |
|---|---|---|
| 9:45 to 10:53 am | A separate Teams incident (TM1550488) in the Americas, blamed on "resource contention within its infrastructure". Microsoft has not said whether it is related. | BleepingComputer |
| 12:39 pm | User reports on Downdetector begin to climb | Techaeris |
| 12:54 pm | Microsoft is investigating Exchange Online mailbox access "through multiple connection methods" in the Americas, incident EX1550543 | Microsoft on X |
| 1:07 pm | Microsoft: the issue "may affect Microsoft 365 service connectivity for users in any region" | Microsoft on X |
| 1:09 pm | Outlook reports on Downdetector peak at about 129,000 (user reports, not confirmed failures) | Techaeris |
| 1:26 pm | The start time in the admin center advisory. BrokenToday records the start as 2:30 pm. | BleepingComputer, BrokenToday |
| 1:35 pm | Microsoft: a portion of infrastructure is under "elevated load across multiple regions" | Microsoft on X |
| 1:48 pm | Microsoft has begun "redistributing connection traffic across our broader worldwide infrastructure"; the post now names EX1550543 and MO1550821 | Microsoft on X |
| 2:12 pm | All deployments paused, more traffic moved off the affected infrastructure, affected servers being restarted | Microsoft on X |
| 2:31 pm | Microsoft: the load is not from more traffic but from "a number of servers being unexpectedly unavailable" | Microsoft on X |
| 2:50 pm | A configuration change is being tested for servers that fail automated health checks; all affected servers are being restarted | Microsoft on X |
| 3:19 pm | No recent change found that matches the start of impact; the failed health checks may come from a component that processes server health | Microsoft on X |
| 4:00 pm | An alternative health check method let a subset of servers return to service and is being applied more broadly | Microsoft on X |
| 4:40 pm | Servers being restarted and validated; Microsoft is "assessing whether reverting a small number of recent configuration updates we identified could help accelerate recovery" | Microsoft on X |
| 5:09 pm | "Opening all available pathways to restore service, as our actions taken thus far have not provided the anticipated relief"; a core process on affected servers is becoming overwhelmed, causing capacity loss | Microsoft on X |
| 6:04 pm | A portion of servers returned to service and stable so far; "limited recovery" | Microsoft on X |
| 7:06 pm | The returned servers "can become unresponsive again once they receive traffic"; some traffic types limited | Microsoft on X |
| 8:59 pm (01:59 UTC, 13 Nov) | "Positive results" restarting infrastructure in an isolated manner, still monitoring. We read this post at about 03:00 UTC; a later copy of the thread did not show it. | Microsoft on X |
| 11:36 pm | Efforts to improve server stability include "adjusting traffic protocols, redistributing traffic, and applying strategic traffic blocking measures"; partial recovery in the regions where applied | Microsoft on X |
| 12:01 am, 13 Nov | "Multiple workstreams to alleviate impact, address our current blocker, and returning portions of service infrastructure back to a healthy state" | Microsoft on X |
| 2:29 am, 13 Nov (07:29 UTC) | Status feed: "positive results for server stability" after applying "an upstream authentication/routing configuration component patch to a portion of our infrastructure"; working to expand it to servers supporting active regions; next update promised by 08:30 UTC | Microsoft status feed |
| 2:44 am, 13 Nov (07:44 UTC) | "Positive results for service stability and increased capacity" from "our latest traffic optimization actions", being implemented "more broadly across affected regions as we seek to attain persistent recovery" | Microsoft on X |
| 4:14 am, 13 Nov (09:14 UTC) | Status feed: "early signs of stabilization, but service capacity has not yet returned to normal levels"; infrastructure cycled through "traffic drain, restart, remediation, and staged return-to-service"; the email timeout window extended to 48 hours | Microsoft status feed |
| 4:24 am, 13 Nov (09:24 UTC) | "Early stabilization and capacity improvements"; the email timeout is extended to 48 hours | Microsoft on X |
| 5:28 am, 13 Nov (10:28 UTC) | "Incremental improvement in several workloads, but service health remains variable and capacity has not yet fully returned to baseline"; the 48-hour email timeout window stays while queue and backlog processing continues. From this post on, Microsoft's X posts name MO1550821 and no longer EX1550543. | Microsoft on X |
| 6:11 am, 13 Nov (11:11 UTC) | Actions "completed across portions of affected infrastructure" and traffic "gradually restored"; improvements in key scenarios, "including mailbox access and delayed email processing" | Microsoft on X |
| 7:09 am, 13 Nov (12:09 UTC) | "Encouraging recovery progress"; the 48-hour email timeout window "remains in place" | Microsoft on X |
| 8:32 am, 13 Nov (13:32 UTC) | "Broader improvements in North America, sustained improvements across much of Europe, and encouraging recovery signals in Asia Pacific" | Microsoft on X |
| 9:17 am, 13 Nov (14:17 UTC) | Email queues "significantly reduced and approaching normal levels"; additional mitigation measures implemented, remaining affected infrastructure being restored, residual service issues under investigation | Microsoft on X |
| 10:39 am, 13 Nov (15:39 UTC) | "For now we can confirm that there is no indication this is the result of malicious or cyber activity"; Microsoft is reviewing what happened and how to prevent similar issues | Microsoft on X |
| 12:57 pm, 13 Nov (17:57 UTC) | Status feed, still "Service degradation": "broad recovery and stability across our services"; "a technical issue not related to cyber or AI / agents"; "introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change"; the underlying issue is addressed and service is still being restored to remaining customers and regions; a full post-incident report is promised; next update promised by 20:00 UTC | Microsoft status feed |
| 1:05 pm, 13 Nov (18:05 UTC) | Same wording as the 17:57 UTC feed post | Microsoft on X |
Teams: ABC News quotes Microsoft saying "mitigative actions have successfully restored service health and functionality" for Teams, with no time given.
The incident numbers: Microsoft's own posts name EX1550543 from the first post and MO1550821 alongside it from 1:48 pm. Microsoft does not say what each covers. BleepingComputer cites MO1550821 and Techaeris cites EX1550543 for the same event, so look for either in your admin center. TM1550488 (the morning Teams incident) appears only in BleepingComputer.
What Microsoft said about the cause
Microsoft's cause statement (13 November, 17:57 UTC feed post, repeated on X at 18:05 UTC): "The incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change. We have addressed the underlying issue and continue to restore service to remaining customers and regions." It adds that this "was a technical issue not related to cyber or AI / agents".
What Microsoft has not said: what exactly failed in the capacity upgrade, which network change it means, whether that is the "recent network configuration change" XDA quoted, or why the servers failed their health checks and why the secure-connection process was overwhelmed earlier in the incident. It promises a full post-incident report once its review is complete; none exists yet. Secondary reports go further than Microsoft's wording and we do not adopt them: BetaNews paraphrases it as "a fault left behind by a past upgrade", and SecurityWeek and others say Microsoft began rolling back a recent network configuration change, which we did not find in the posts we read (single source, unverified).
The earlier statements, as the outlets quoted them and as Microsoft posted them during the incident:
- First: "a portion of infrastructure is experiencing elevated load across multiple regions", which Microsoft said was "likely contributing to the connectivity errors", and it was reviewing "an unexpected increase in a specific type of service traffic".
- Then: "We determined the elevated load isn't driven by an increase in traffic". The cause was "a number of servers being unexpectedly unavailable", and Microsoft said it was working "to safely return them to service".
- Mitigation, as reported: all deployments paused, connection traffic redistributed worldwide, traffic moved off the affected systems, affected servers restarted. XDA also reports that configuration changes were paused while Microsoft reviewed recent changes.
- Microsoft's post at 3:19 pm ET said it had not identified a recent change correlated with the start of impact. At 4:40 pm it said it was assessing whether reverting "a small number of recent configuration updates we identified" could speed recovery. XDA quotes a later update saying it had begun reverting "a recent network configuration change that correlates with the start of impact"; we did not find that exact post in the thread. Microsoft gave its cause statement only later, as quoted above.
Our reasoning, not Microsoft's: the pattern of some servers dropping out and the rest being overloaded is a cascade. Fewer healthy servers means more load on each, which makes more of them fail their health checks, so restarting a few at a time is slow by design. That would explain why restoring service was reported as taking hours, but it is an inference from the wording, and Microsoft's own explanation (a network capacity upgrade and a recent network change) does not say how that led to servers dropping out.
What to do now
If you run Microsoft 365 for a company or customers:
- Confirm it is Microsoft, not you. Look at Service health in the Microsoft 365 admin center for the incident number, and at the @MSFT365Status account on X. If the admin center will not load, that is part of the symptom. The public page is status.cloud.microsoft.
- Do not change anything on your side. Our reasoning: password resets, re-creating profiles, changing MX records or rebuilding Outlook profiles will not fix a server-side outage, and they add new problems that you have to clean up later.
- Tell your users early. One message with the incident number and "no action needed" prevents a flood of tickets. If everyone's mail and chat are down, say where updates will come from (a phone message, an SMS or a status page that is not on Microsoft).
- Expect delayed mail, not lost mail. General SMTP behavior, not a Microsoft statement: a sending server that cannot deliver normally retries for days, so mail sent to you during the outage should arrive late. Only after the incident is closed, check for bounces that your own senders received and for meetings that were created or deleted during the window. Boston University's page on a similar incident in November 2024 reports deleted events reappearing and new events disappearing, which is the kind of thing to look for.
- Delayed mail has a longer window. From Microsoft's posts: it extended the email timeout window to 48 hours, so expect queued mail to be delivered, and its 9:17 am ET (14:17 UTC) post says email queues are approaching normal levels. In the admin center, check Service health for MO1550821, the incident Microsoft's posts on X have named since 10:28 UTC on 13 November. We cannot see the per-incident end state, because the admin center needs a login.
- Keep the incident number and the times for your records and for any service credit request. Whether you qualify depends on your agreement.
Quick "is it us?" checks from a machine outside your network. Standard, read-only checks, covered in the guides linked below:
# Does your domain still point at Exchange Online? (read-only)
dig +short MX example.com
# Does the service answer at all from outside? (read-only)
curl -sI https://outlook.office.com | head -n 1If the MX record is what you expect and the HTTPS request answers, your side is fine and the problem is a sign-in or mailbox-access failure inside Microsoft's service, which matches what Microsoft described. On RHEL, AlmaLinux and Rocky, dig comes from the bind-utils package.
The lesson for your monitoring
This part is our reasoning, not Microsoft's.
Many teams get their email, their chat and their ticket alerts from the same Microsoft tenant. When it goes down, the alert that tells you about it can be stuck in the same outage. A monitoring alert that is emailed to a Microsoft 365 mailbox does not arrive while Exchange Online is down.
A small independent setup costs little:
- Alerts to a second path. At least one alert target that does not use your Microsoft 365 tenant: a phone number, a personal mailbox elsewhere or a chat channel on another platform.
- An outside check of your own services. An HTTP and port check from the internet for your public site, VPN gateway and mail endpoint, so you know if the problem is yours or Microsoft's.
- A heartbeat for backups and jobs. The job reports to a separate service when it finishes, and silence raises the alert. That works even when your mail is down.
- A written "how do we talk" plan. Where does the team go when Teams and Outlook are both down? Decide before the outage, not during it.
None of this replaces Microsoft's status page. It makes sure you hear about trouble from your own checks, in a place Microsoft cannot take down.
What we have not verified
- Whether the incident is over, and the final time. Microsoft's status feed post authored 17:57 UTC on 13 November still says "Service degradation", and none of its feed posts or posts on X says resolved. The per-incident end state is not public (the admin center needs a login). BleepingComputer (updated 17:50 UTC) says the separate Teams incident TM1550488 was fully resolved; that is a different ID and a secondary source. We do not use incident-tracker records for the end time, and we do not use the times press outlets give, because several conflict with Microsoft's own timestamps.
- Whether the cause statement is complete. Microsoft says an issue during an earlier network capacity upgrade surfaced after a recent network change. We do not know which change, and Microsoft's earlier posts said no recent change correlated with the start of impact. The 07:29 UTC "upstream authentication/routing configuration component patch" is described by Microsoft as a mitigation, not a cause.
- Whether a rollback happened. SecurityWeek and others say Microsoft began rolling back a recent network configuration change; we did not find that in Microsoft's posts (single source, unverified).
- Press paraphrases. BetaNews's "a fault left behind by a past upgrade" is not Microsoft's wording. ABC News and BleepingComputer repeat Microsoft's no-cyber statement and the 14:17 UTC post respectively; they are secondary sources.
- Microsoft's update commitments: the feed promised a next update by 08:30 UTC and, in the 17:57 UTC post, by 20:00 UTC. We checked at about 19:50 UTC and did not see whether either was met.
- The post-incident report. Microsoft says it will publish a full report once its review is complete; none exists yet.
- Hybrid mail flow: a third-party mirror of Microsoft's Service Health text for EX1550543 (cloudadminhub.com, single source, not Microsoft's own page) says "on-prem mail flow may be affected if it goes to or from Exchange Online". We did not see that sentence in Microsoft's public status feed.
- The mechanism. "Servers unexpectedly unavailable" and "a core process becoming overwhelmed" describe the effect; Microsoft's cause statement names the upgrade and the change but not how they produced it.
- What the two incident numbers (EX1550543 and MO1550821) each cover, and whether the morning Teams incident (TM1550488) is related. Microsoft has not said.
Related guides
- Test connectivity with ping and curl: the outside check that shows whether a service answers at all.
- Diagnose Linux DNS with getent and dig: reading an MX or A record and telling a DNS problem from a service problem.
- Monitor a restic backup from a systemd timer: a heartbeat that alerts on silence, with no dependence on your mail.
- Backup monitoring with heartbeat checks: the dead man's switch pattern in detail.
- N-able N-central offline: what MSPs should do now: the same lesson from this week's other outage, a management tool that went down with the alerts inside it.
Sources
- BleepingComputer, "Massive Microsoft 365 outage disrupts Outlook, Teams, Copilot worldwide": https://www.bleepingcomputer.com/news/microsoft/massive-microsoft-365-outage-disrupts-outlook-teams-copilot-worldwide/
- Techaeris, "Microsoft 365 Outage Takes Down Outlook, Teams and Copilot": https://techaeris.com/2026/11/12/microsoft-365-outage-takes-down-outlook/
- Engadget, "Multiple Microsoft 365 services are experiencing an outage": https://www.engadget.com/2315259/multiple-microsoft-365-services-are-experiencing-an-outage/
- XDA Developers, "Microsoft is suffering a major outage, taking down Microsoft 365, Outlook, Teams, OneDrive, and more": https://www.xda-developers.com/microsoft-suffering-major-outage-taking-down-microsoft-365-outlook-teams-onedrive/
- Microsoft 365 Status on X, the incident thread (read 13 November 2026, about 19:45 UTC; latest post 18:05 UTC): https://x.com/MSFT365Status/status/2121002505911857577
- Microsoft service status feed (newest post authored 17:57 UTC on 13 November, read at about 19:45 UTC): https://status.cloud.microsoft/
- ABC News, "Microsoft Outlook outage stretches on for hours" (secondary; published 13 November, quotes the no-cyber statement): https://abcnews.com/US/outlook-down-microsoft-outage-reported-mail/story?id=138088203
- BleepingComputer, 13 November update to the article above (secondary; quotes the 14:17 UTC post, says TM1550488 was resolved).
- ynetnews, "Microsoft services back online after global outage": https://www.ynetnews.com/tech-and-digital/article/hy96sy70mx
- BrokenToday, an incident tracker (secondary source; we use it only for the start time): https://brokentoday.com/is-microsoft-365-down/
- Boston University IT, the November 2024 Outlook and Teams incident (for the calendar side effects): https://www.bu.edu/tech/?p=155033
- Microsoft service status: https://status.cloud.microsoft/
More from the blog
- wfx GitHub Actions Stealer: Check Repos, Cloud Roles · News, 2026-11-12
- N-able N-central Offline: What MSPs Should Do Now · News, 2026-11-10
- Appwrite Incident 1 Nov 2026: Rotate Secrets · News, 2026-11-07
This article is educational information, not incident-response or legal advice. Verify package status with your operating system, hosting, or runtime vendor.