Microsoft works to restore Outlook and Microsoft 365 Copilot services
Microsoft was working to restore Outlook and Microsoft 365 Copilot after an outage disrupted Exchange Online-dependent features and delayed some email delivery.
Microsoft was working to restore Outlook and Microsoft 365 Copilot after a Microsoft 365 service incident disrupted email access, delayed some delivery and impaired AI features. The incident began at 18:26 UTC on November 12. In the latest opened update, issued at 05:43 UTC on November 13, Microsoft said recovery work was continuing.
Exchange Online users could be unable to reach mailboxes through Outlook on the web, Outlook desktop using MAPI, Exchange ActiveSync and REST-based connections. Microsoft said email delivery could also be delayed. Microsoft 365 Copilot users could be unable to load Copilot Chat or receive responses in web and desktop apps. Copilot experiences could also lose access to work data including email, calendars and files.
Microsoft narrowed the impact to features that interact with Exchange Online and email services. Parts of the listed services that did not depend on Exchange Online remained functional.
The company said a process that establishes secure connections on affected servers had become overwhelmed. That caused servers to stop processing requests and reduced available capacity. Microsoft described the mechanism while its investigation was still underway, not as a final root-cause finding, and said it had found no indication of malicious activity.
Recovery efforts included restarting servers, redirecting traffic, isolating parts of the infrastructure, rebuilding a subset of servers with an earlier configuration and applying capacity changes. Microsoft said some mitigations did not provide the expected relief or were not taking effect as expected.
Microsoft reported limited recovery in some regions at 22:58 UTC as clients reconnected. However, the latest opened update did not establish full resolution or give a resolution time. The company said it would publish a post-incident report after mitigation and scheduled its next service-health update for 07:00 UTC on November 13.
More news

Microsoft and NVIDIA publish open-weight model risk framework

Microsoft broadens Sovereign Cloud controls across three deployment models

Microsoft disrupts EvilTokens, citing 12,000 compromised inboxes
