Microsoft’s global 365 outage stretched past 13 hours with the company reporting only “marginal success” from its latest recovery measures and no clear timeline for full restoration. In Australia, where the disruption hit the country’s spy agency and federal police, the government triggered an agreement with Microsoft for priority restoration of services. Microsoft reiterated that it has found no indication of malicious or cyber activity and said it will file a post-incident report once the disruption is resolved.
The breakdown began at 16:17 UTC Thursday on Exchange Online after a batch of servers unexpectedly went offline and overwhelmed a core process on remaining infrastructure. Traffic rerouting, protocol adjustments and server restarts produced only partial recovery because restored servers repeatedly became unresponsive again once traffic returned, even as Microsoft rebuilt a subset of servers using a previous configuration and applied capacity upgrades. The Verge’s Tom Warren said a rollout spiked server CPU usage, forcing racks into maintenance mode and triggering cascading failures, while commentators noted the incident followed Microsoft’s Oct 26 buyout offer to about 60,000 employees.
Problem reports on Downdetector spread to 74 countries and territories within an hour and reached 19,650 per minute across five Microsoft services, compared with fewer than 15 on an average Thursday. Outlook reports peaked at more than 2,000 times normal levels and topped 130,000, alongside disruptions to Teams, Copilot, OneDrive, SharePoint, the Defender XDR security platform and Microsoft connectors in OpenAI and Anthropic tools. The outage also cut off email access for a US state government, two Michigan hospital systems and Indian airline IndiGo’s customer support.
