Microsoft's Fixes Weren't Working: Inside the 12-Hour Global 365 Outage

Microsoft's attempts to restore its own services were failing in public. More than 12 hours after a global Microsoft 365 outage began on Thursday, November 13, 2026, the company reported only "marginal success" in bringing Outlook, Teams, and Copilot back online for users in 74 countries.

The problems started around 9:30 a.m. when users began reporting issues on Downdetector. Reports peaked at nearly 20,000 per minute across five Microsoft platforms. Microsoft said the outage was not caused by a cybersecurity incident. Its recovery efforts told a different story: some restored servers could become "unresponsive once again" when they received traffic, and changes made to improve the situation were "not taking as expected."

The scale is hard to overstate. More than 430 million people use Microsoft 365 apps. The disruption hit email, calendars, and AI tools for businesses, governments, and healthcare systems. UK Parliament said it was affected and had "alternative communication arrangements" in place. NHS GP practices reported disruption, with one doctor describing Microsoft services as "central" to their work. In Australia, spy and federal police agencies reported issues.

Microsoft did not provide a timeline for full restoration. When a single provider carries this much of the world's productivity load, a failed recovery is not just an outage — it is a concentration risk made visible. Governments and critical services had no clear path back to normal because the company could not say when normal would return.

The financial and productivity costs are not yet quantified. What is established: the world's productivity backbone went down, and its owner could not reliably bring it back on demand.