Monday–Friday, 9:00 AM–5:00 PM AEST · After-hours support available on request 1300 584 007
Microsoft 365 global outage November 2026 incident timeline graphic

If Outlook, Teams or Copilot stopped working for you in the middle of last week and you assumed it was just your internet, it wasn’t. On 12–13 November 2026, Microsoft had one of its most significant Microsoft 365 outages in years, affecting Exchange Online, Outlook, Teams, Copilot, SharePoint Online and more than a dozen other services, for businesses around the world — Australia included. If you’d rather not work through the next one alone, that’s what our Microsoft 365 support is for.

In short: a growing number of Exchange Online servers began failing Microsoft’s automated health checks and dropping out of service. As they dropped out, the remaining servers were overloaded and began failing too, in a cascade that spread to Outlook, Teams, Copilot and more than a dozen dependent services. Microsoft says it wasn’t a cyberattack. Full recovery took roughly a day and a half of staged, region-by-region restoration.

When did it actually start?

Microsoft tracked this as two linked incidents, and posted updates through the day on its Microsoft 365 Status account. The first, EX1550543 (Exchange Online mailbox access), began around 8:17am Pacific time on 12 November — about 4:17am AEDT on 13 November for Australian readers. It widened into a broader Microsoft 365 outage, tracked as MO1550821, by around 10:26am Pacific (6:26am AEDT), at which point Teams, Copilot, SharePoint Online, Intune, Defender for Office 365, Planner, To Do, Purview, Power BI, the Microsoft 365 Apps admin center, Universal Print and Cloud PCs were all added to the list of affected services over the following hours.

What actually went wrong?

Microsoft’s own incident updates, posted roughly every 30–60 minutes throughout the day, describe a specific and somewhat unusual failure pattern: a growing number of servers began failing their own automated health checks and were pulled out of service as a precaution. But the load those servers had been handling didn’t disappear — it landed on the remaining, still-healthy servers, which then started failing under the extra strain too. Microsoft said the elevated load “isn’t driven by an increase in traffic, but by a number of servers being unexpectedly unavailable.”

For much of the day, Microsoft’s engineers couldn’t pin down what had triggered it. Early updates said plainly that no recent change had been found that correlated with when the impact started. By the evening (US time), Microsoft attributed the trigger to “an issue during an earlier network capacity upgrade that surfaced after a recent network change” — language that points to a delayed side-effect of earlier infrastructure work, rather than a single obvious mistake made that day.

Microsoft was also clear and consistent on two points throughout: this was not caused by a cyberattack, and it had nothing to do with AI systems or Copilot’s underlying models. Several updates stated directly there was “no indication this is the result of malicious activity.”

How did Microsoft fix it?

There was no single fix. Microsoft’s incident updates, posted on its Microsoft 365 Status feed and in the admin centre’s Service health page, show a long run of attempts, several of which didn’t hold the first time. The times below are approximate and converted to Sydney time.

When (Sydney time) What Microsoft did
From about 5:30am, Friday 13 November Redistributed traffic worldwide, paused all deployments and configuration changes, and began restarting affected servers.
About 7am Deployed a targeted change to bypass a component on some servers, then restarted large numbers of unresponsive servers.
About 8am to 9am Validated an alternative health-check method on some servers and applied it more widely, then isolated part of the infrastructure to test recovery.
About 11am Servers that had recovered became unresponsive again under traffic. Microsoft limited some traffic types and rebuilt servers on a previous configuration.
About midday Began reverting a recent network configuration change, and reported “positive results” from restarting infrastructure.
Friday night into Saturday 14 November Restored service in stages, checking stability at each step, with recovery reported across North America, Europe and Asia Pacific and email queues clearing.

When were services actually working again?

Recovery was gradual, not a single “all clear” moment. Microsoft said it had addressed the underlying issue and was restoring service to the remaining customers and regions in stages, which took longer but was meant to avoid the repeat failures seen earlier, when recovered servers buckled again under returning traffic. On Saturday 14 November (US time) Microsoft said services were fully restored, after an outage that ran across two working days.

Has Microsoft published a full explanation yet?

Not yet, as far as we can confirm. Microsoft has promised a full post-incident report once its review is finished. Until then, the explanation above (the earlier network capacity upgrade, the servers failing health checks and the overload that followed) is Microsoft’s preliminary account, not the final word.

A fallback plan sits alongside your backup and disaster recovery planning: one covers losing data, the other covers losing access.

FAQs

Was my data at risk during this outage? Nothing reported publicly suggests data loss or a security breach — this was described throughout as an availability problem (services being unreachable), not a data-integrity or confidentiality incident.

Why did some people in my office lose Outlook while others didn’t? The outage affected specific infrastructure regions and components rather than every server globally at once, which is consistent with why impact was patchy — some users, regions or specific features (like Exchange-dependent features specifically) were hit harder than others.

Could this happen again? Any large cloud provider can have an outage; what matters more than the risk of zero is how a business is set up to keep working through one — a documented fallback for email and communication is the practical takeaway here, not avoiding cloud services altogether.

Is this related to the other recent Microsoft issues Syntrex has covered, like the stolen M365 vendor credential incident? No — that was a separate security incident involving compromised login credentials at a third-party vendor. This outage was an infrastructure availability issue, with Microsoft saying explicitly it found no evidence of malicious activity.

The takeaway

An outage of this size and duration at a provider as large as Microsoft is a reminder that “the cloud” is still physical infrastructure that can fail, however reliable it usually is. If your business doesn’t have a simple fallback for critical communication — a way to reach staff or customers if Outlook or Teams genuinely isn’t available for a day — it’s worth thinking through before the next one, rather than during it. If you’d like help putting a practical continuity plan together, call us on 1300 584 007.

Outages are rare. The more common Microsoft 365 risk is covered in our post on why stolen logins are now the main way attackers get in.

Not sure where your business stands? Book a free IT consultation.

Need a hand with this?

Get in touch and we'll walk you through it.