Major Microsoft 365 outage causes authentication and email failures in Exchange Online

Microsoft confirmed a far-reaching Microsoft 365 incident that caused authentication problems, connection delays and service failures for numerous customers. The company first acknowledged the issue at 17:30 UTC on August 31, 2026, initially logging it under the identifier EX1464935 in the admin center, after user complaints about Exchange Online began multiplying on social media. Although this was the most affected service, Microsoft later acknowledged, in an incident report renumbered as MO1465074, that the failure also reached OneDrive for Business, SharePoint Online, Microsoft Teams, Microsoft Purview and Microsoft Defender XDR.

According to the monitoring service Downdetector, tens of thousands of Outlook and Microsoft 365 users reported problems, mainly related to receiving email in Outlook and accessing Microsoft's cloud services. The company itself detailed a list of symptoms associated with Exchange Online: delays or failures sending and receiving email messages; delays, failures or incomplete results when searching content within mailboxes; authentication errors when accessing the services; difficulties accessing or performing actions in Exchange administration tools; and intermittent failures in mailbox operations and message delivery flows. Microsoft classified the event as an "incident," a category usually reserved for critical problems with notable user impact, though it did not specify which geographic regions were affected.

In an update, Microsoft explained that it had "isolated a common failure pattern in the affected Exchange Online requests, associated with authentication and protocol connectivity," and that it was analyzing service telemetry to identify the source of the problem and validate possible fixes. Later, the company was more specific and said it had identified "issues related to a basic authentication configuration used by multiple internal services within the Exchange Online infrastructure," and was therefore running manual tests at the individual server level to reset configurations and check whether that resolved the failure.

The update timeline published by Microsoft shows a gradual recovery, and not one free of back-and-forth. At 15:09 EDT on August 31, the company began tracking the incident under the new identifier MO1465074, confirming its scope across the additional services already mentioned; shortly afterward, at 16:12 EDT, it revised the advisory title to reflect that the problem affected multiple Microsoft 365 services and not only Exchange Online. Twenty-two hours after the problem was first acknowledged, in the 10:29 EDT update on September 1, Microsoft reported "positive recovery trends" in service telemetry, indicating that it was continuing to apply a targeted fix to the affected infrastructure and validating that authentication, connectivity and searches were returning to normal levels. At 12:02 EDT that same day, the company announced that mail flow and search functionality in Exchange Online had been restored, and that the focus was shifting to validating recovery across the remaining affected Microsoft 365 services.

However, the problem was not fully resolved right away: in the most recent update cited in the article, dated September 2 at 12:00 EDT, Microsoft acknowledged that, according to its telemetry, "a small subset of users" might still be experiencing impact in the scenarios noted, while the company continued to balance the rollout of the fixes across the infrastructure to achieve a durable recovery.

The article frames this episode within a series of recent Exchange Online incidents. In June, Microsoft had already mitigated a widespread problem affecting the mail flow pipeline for customers in North America, Asia-Pacific and Europe. In April, the company also resolved mailbox access problems that had intermittently affected Outlook mobile and macOS users for weeks, as well as another Exchange Online outage that blocked mailbox access through Outlook on the web, Outlook desktop, Exchange ActiveSync and other connection protocols.

In reader comments collected in the article, some system administrators recounted their own experiences during the incident: it was noted that restarting Outlook worsened connection problems, while keeping the application open allowed work to continue normally, and that one user was successfully authenticated by repeatedly launching Outlook in safe mode, although that method later stopped working as well. These accounts are user reports in the comments, not official Microsoft confirmations, but they illustrate the intermittent and uneven way the failure affected different accounts and configurations throughout the recovery process.