Overview
- Users reported large-scale problems sending, receiving and opening email and other Microsoft 365 tools after error reports spiked on Thursday and Downdetector recorded roughly 120,000–140,000 reports at the peak.
- Microsoft says the problem was not higher user traffic but a subset of servers becoming unexpectedly unavailable because they failed automated health checks, which shifted load onto the remaining infrastructure.
- In response, Microsoft paused all deployments, redistributed connection traffic across global infrastructure, restarted affected servers and tested an alternative health-check method that let some machines return to service.
- The outage disrupted businesses, government offices, schools and third-party AI connectors such as OpenAI and Anthropic, and administrators were directed to follow incident EX1550543 and MO1550821 in the Microsoft 365 admin center.
- Microsoft reports no evidence of a cyberattack so far and is preparing automated recovery systems while it continues to investigate why the server health checks and processor usage failed and what broader fixes are needed.