Microsoft 365 outage traced to latent cloud gateway defect
A dormant bug in Microsoft's cloud gateway caused a 24-hour global outage affecting 74 countries and critical public services.
Microsoft 365 experienced a major outage from November 12 at 16:20 UTC to November 13 at 16:10 UTC, impacting 74 countries. A preliminary Post-Incident Report identified the cause as a technical issue unrelated to cyber threats or AI agents. The incident stemmed from a hidden defect in the cloud gateway, dormant since July, which exhausted CPU and memory resources under specific conditions. This latent bug escaped pre-release testing because the hardware used was less powerful than production systems. Standard network maintenance shifted Teams traffic to the primary gateway, exposing the defect and causing more than 80 percent of affected servers to stop responding.
The disruption left businesses unable to process transactions and public sector workers struggling to function. At least 47 NHS trusts reported problems, with some services reverting to pen and paper. Ian Murray, minister for creative industries, media and online safety, acknowledged that 98% to 99% of civil servants use Microsoft for communications. The public sector spent £1.9 billion on Microsoft licenses in 2024-25. While the government intends to use collective procurement to support British businesses, it offered no timetable for reducing reliance on Microsoft or adopting open-source alternatives.
Sources
- Ars TechnicaThree small failures combined to take down Microsoft 365 ↗
- The RegisterUp to 99% of UK civil servants rely on Microsoft for comms, minister admits ↗
Every sentence of this story was checked against facts quoted from these articles. How we make the news
Spotted an error? Tell us
