Overview
- Microsoft said the outage began on Thursday and lasted almost 24 hours as cascading failures made many gateway servers unresponsive during U.S. peak traffic.
- The company identified three linked causes: a latent software defect introduced by a July patch, an erroneous Teams failover about 15 hours earlier that shifted load to the primary gateway, and roughly 14% less gateway capacity from an unrelated issue the day before.
- Engineers developed and deployed a corrective update that reverted the problematic code path and used a staged recovery process to bring infrastructure back without immediately overwhelming servers.
- Core Microsoft 365 services — including Exchange Online/Outlook, Teams, Copilot, OneDrive, SharePoint and Office apps — were disrupted for millions of users worldwide and outage trackers recorded sharp spikes in user reports.
- Microsoft said it found no signs of malicious activity or AI failure, published preliminary incident records under MO1550821 and EX1550543, and warned the episode will prompt further technical review when the final post‑incident report appears this week.