When the Cloud Goes Quiet: Inside Microsoft’s 24-Hour Microsoft 365 Outage
On 12–13 November 2026, Microsoft 365 went down for nearly a full day across 74 countries. Email, Teams, SharePoint, Copilot and even Microsoft’s own security tools stopped working for millions of organisations. It was not a cyberattack. It was a failure inside Microsoft’s own infrastructure, triggered by a network change. This article explains what happened, why the fixes kept failing, and what businesses should learn about availability and cloud dependency.
Ilie Lucian - Founder & CyberSecurity Engineer
11/13/20265 min read


Introduction: when email stops, everything stops
Most organisations treat email as something that simply works. On Thursday 12 November 2026, it stopped working for a large part of the world.
Starting in the afternoon in Europe, users began losing access to their Exchange Online mailboxes. Within hours, the problem spread to Outlook, Teams, SharePoint, OneDrive and Microsoft 365 Copilot. Downdetector showed disruption to Microsoft products in 74 countries and territories. In the US alone, it logged more than 130,000 Outlook reports at the peak.
The outage lasted almost 24 hours. On Friday, Microsoft apologised: “We are deeply sorry for the disruption this incident has caused our customers and partners.”
Timeline
Microsoft tracked the incident as EX1550543 (Exchange Online) and MO1550821 (wider Microsoft 365). Times below are in UTC.
12 Nov, 16:17: Users begin reporting that they cannot open their mailboxes.
12 Nov, ~18:00: Reports spike worldwide. The Register dates the main outage from around this time.
12 Nov, 18:19: Microsoft reports elevated load across multiple regions and begins restarting servers and redirecting traffic.
12 Nov, ~19:06: Microsoft pauses deployments and configuration changes across the service.
12 Nov, 19:43: Affected servers are confirmed to be failing health checks and are being pulled from service.
12 Nov, 20:56: An alternative health-check method brings partial recovery.
13 Nov, ~00:00: Servers that were returned to service become unresponsive again under production traffic.
13 Nov, before 01:50: Microsoft starts reverting a recent network configuration change that correlated with the start of impact.
13 Nov, 03:07: Microsoft says its changes are “not taking as expected.”
13 Nov, 13:00: Recovery is visible in North America and much of Europe, with “encouraging” signs in Asia Pacific.
13 Nov, afternoon: Microsoft reports “broad recovery and stability” and apologises publicly.
What actually broke
Microsoft’s status updates describe a classic cascading failure.
1. Servers failed their health checks. Cloud services constantly test their own servers. If a server fails an automated health check, the system removes it from service so users are not sent to a broken machine. On 12 November, a large number of Exchange Online servers began failing these checks at the same time.
2. The remaining servers were overloaded. Microsoft stressed that overall traffic had not increased. The problem was that so many servers dropped out that their work landed on the ones still running. Those servers then struggled too.
3. A core process was overwhelmed. In a later update, Microsoft said that “a process on the affected servers responsible for establishing secure connections becomes overwhelmed”. Once that happened, servers stopped handling requests, and more capacity was lost.
4. Recovery kept collapsing. Every time engineers returned servers to service, live traffic hit them again and they failed. Attempts to add capacity by scaling virtual machines ran into machines stuck in a bad state. As Microsoft admitted overnight: “Actions taken thus far have not provided the anticipated relief.”
Who is to blame?
Microsoft itself. According to its preliminary explanation, the incident was “introduced by an issue during an earlier network capacity upgrade” that only surfaced after a more recent network change. In other words, a latent problem sat in the network for some time, and a later change triggered it.
Some details are still unclear. The Verge’s Tom Warren linked the outage to Azure virtual network configuration changes, but Microsoft has not confirmed this. A full root-cause analysis has been promised in a post-incident report, which has not been published yet.
Was it a cyberattack?
No evidence suggests it was. Overnight, Microsoft said there was “no indication this is the result of malicious activity”. It later stated that the problem was not related to cybersecurity or AI. Australia’s Home Affairs Minister Tony Burke compared it to the CrowdStrike incident of July 2024 and said Microsoft described it as resembling an update gone wrong.
The comparison is useful but imperfect. CrowdStrike’s failure came from a faulty third-party security update that crashed around 8.5 million Windows devices. This time, no customer devices were damaged. The failure was entirely inside Microsoft’s cloud.
The impact
The list of affected organisations shows how deeply Microsoft 365 is built into daily work:
Public sector: the UK’s NHS and Metropolitan Police, Immigration New Zealand (which paused visa processing), Australian federal agencies and the US House of Representatives.
Business and finance: Euronext, Aviva, Maersk and Richemont. Workhorse Group even postponed its quarterly results.
Airlines: IndiGo, Akasa Air and Air India reported problems with email support.
Japan: the government’s shared IT environment and around 7,500 staff in Mie Prefecture were affected.
Emails that could not be delivered were held in queues until at least Saturday afternoon before being returned as undeliverable.
The security angle: availability is part of security
Security is usually discussed in terms of attackers. But the classic CIA triad, Confidentiality, Integrity and Availability, treats availability as a core security goal. This outage is a lesson in that third pillar.
Three points stand out for security teams:
Security tools went down too. Microsoft’s status page listed Defender XDR and Purview among the affected services, along with Safe Links. Organisations that rely only on Microsoft for email protection and monitoring lost part of their visibility during the outage.
Outages are phishing season. When users cannot reach their email, they become easy targets for fake “Microsoft 365 service restoration” messages and support scams. Note the timing: Microsoft had patched an exploited Outlook spoofing flaw, CVE-2026-124194, only two days earlier. Microsoft has not linked it to the outage.
Concentration risk is real. When email, chat, file storage, AI tools and security monitoring all sit with one provider, one failure takes all of them down at once.
What organisations should do
During an outage
Check the official source first. Use Service health in the Microsoft 365 admin centre, or status.cloud.microsoft if you cannot sign in.
Do not “fix” what is not broken. Tell users not to reinstall Office, delete Outlook profiles or reset passwords. None of this helps with a service-side outage.
Freeze your own changes, so that any new problem after recovery is easy to trace.
Warn staff about phishing that uses the outage as bait.
Before the next one
Agree on an out-of-band communication channel, such as Signal, WhatsApp, SMS or a phone tree, that does not depend on Microsoft 365.
Send Service health alerts to at least one address outside Microsoft 365. If your only alert inbox is in Exchange Online, you will not get the alert.
Keep critical contacts and procedures offline, including your incident response plan.
Assess concentration risk. For regulated sectors, this is no longer optional. Under the EU’s DORA and NIS2 rules, dependence on critical ICT providers must be identified and managed.
Save the post-incident report when Microsoft publishes it. Service health keeps history for only 30 days, and you may need the record for SLA claims or audits.
Conclusion
The 12–13 November outage was not caused by hackers, ransomware or a zero-day. It was caused by a network change inside the world’s most widely used productivity cloud, and it still stopped governments, hospitals, airlines and businesses for a full day.
The lesson for every organisation is simple: the cloud is someone else’s infrastructure, and it will fail eventually. Security planning has to include the day your provider goes quiet.
Sources
The Register (2026, Nov 13). Microsoft’s massive M365 outage stretches past 20 hours as its fixes keep failing. https://www.theregister.com/saas/2026/11/13/microsofts-massive-m365-outage-stretches-past-20-hours-as-its-fixes-keep-failing/5309609
XDA Developers (2026, Nov 13). Microsoft’s major outage drags on, leaving Outlook, Teams, OneDrive, and other services still down. https://www.xda-developers.com/microsoft-suffering-major-outage-taking-down-microsoft-365-outlook-teams-onedrive/
SecurityWeek (2026, Nov 12). Microsoft 365 Outage Disrupts Outlook, Exchange Online and Teams. https://www.securityweek.com/microsoft-365-outage-disrupts-outlook-exchange-online-and-teams/
Dawn / Reuters (2026, Nov 13). Microsoft ‘deeply sorry’ as outage triggers global chaos. https://www.dawn.com/news/2044258/microsoft-deeply-sorry-as-outage-triggers-global-chaos
BigGo Finance (2026, Nov 13). Microsoft’s Nearly 24-Hour 365 Outage Revives CrowdStrike Fears, but the Flaw Is Different. https://finance.biggo.com/news/84208d3b-84a8-4ca8-94ff-4a644240d77f
Azure Blog UK (2026, Nov). Microsoft 365 outage, 12 November 2026: what happened and what admins can do. https://azureblog.co.uk/posts/microsoft-365-outage-12-november-2026/
Microsoft Q&A (2026, Nov). Microsoft outage November 2026. https://learn.microsoft.com/en-us/answers/questions/6069789/microsoft-outage-november-2026
1Cyber.eu
Cybersecurity and NIS2 compliance for businesses in Cyprus
© 2026 1Cyber
