Microsoft 365 Outage Hits Outlook, Teams and Copilot in Americas

Microsoft 365 went dark across the Americas on Thursday afternoon, and Copilot went down with it. CNBC reported that Outlook, Teams, Exchange Online, SharePoint Online, Microsoft 365 Copilot and Defender XDR all lost connectivity, with Downdetector logging roughly 128,000 Outlook error reports at the peak.

The company's own words on the cause, from the admin-center notice tracked as EX1550543 and MO1550821: the trouble came from a "number of servers being unexpectedly unavailable, straining the remaining infrastructure." Affected machines were failing automated health checks. Microsoft said the failures "may come from a specific component on the servers that processes those checks," and that a workaround bypassing that component let a subset of servers return to service.

Nothing in the notices pointed to a cyberattack. Microsoft acknowledged the incident at 12:54 p.m. ET, and hours later said it was seeing only limited recovery.

The AI piece was downstream. BleepingComputer reported that Copilot and competing AI services were "unable to access users' inboxes because the underlying Microsoft services were unavailable." The assistants kept running; the data they needed was gone. This is our 78th Microsoft story in the last 90 days, and most of them have been about the AI Microsoft is selling rather than the control plane underneath it.

The notices still do not explain why the health checks failed in the first place, or why processor usage rose on the affected servers.