In this article
  1. What happened
  2. Who’s affected
  3. What to do now
  4. Updates

Developing story, updated as details confirm.

What happened

Microsoft’s Exchange Online incident EX1550543 began at 16:17 UTC on November 12 (12:17 a.m. November 13, Manila), with the wider Microsoft 365 advisory MO1550821. In its latest post, at 22:00 UTC November 13 (6:00 a.m. November 14, Manila), Microsoft’s official service health feed says it continues “to observe sustained service recovery” and is monitoring “for an extended period,” including the backlog of email accrued during the event. The feed still labels it a service degradation. Its next update was due by 02:30 UTC (10:30 a.m. Manila) but had not appeared when we checked at 10:55 a.m.

Microsoft has said it has “no evidence of malicious activity” and will publish a Post-Incident Report. Anthropic marked its Microsoft 365 connector incident on status.claude.com resolved at 5:49 a.m. Manila on November 14, but warns of “occasional errors or delays” while Microsoft completes recovery.

Who’s affected

Exchange Online users on Outlook on the web, Outlook desktop (MAPI), Exchange ActiveSync or REST connections. Microsoft also lists Teams calendar features, Copilot, Graph, Purview, Intune and other services that depend on Exchange.

What to do now

  1. Expect delayed mail to keep arriving; check before resending.
  2. Keep another channel open for customers until access is steady.
  3. Wait for Microsoft’s next update and the Post-Incident Report before changing mailbox or connector settings.

See our guide to Claude and Codex usage limits. More coverage is on our news page.

Updates

  • Nov 14, 2026, 10:55 a.m. Manila (UTC+8): Microsoft’s 22:00 UTC post reports sustained recovery and extended monitoring; its update due by 02:30 UTC had not posted yet. Anthropic resolved its Microsoft 365 connector incident at 21:49 UTC. Headline, summary and advice updated from the official feed.
  • Nov 13, 2026, 10:35 p.m. Manila (UTC+8): Microsoft’s 14:01 UTC update moves EX1550543 to monitoring, with queues near normal, some delays, no evidence of malicious activity and a Post-Incident Report promised. Headline, summary and advice updated.
  • Nov 13, 2026, 12:27 p.m. Manila (UTC+8): Initial report, from @MSFT365Status posts, status.cloud.microsoft and Anthropic’s status page.

Sources & method

Last checked: Nov 14, 2026, 10:55 a.m. Manila (UTC+8). Method: We read Microsoft’s official service health RSS feed, status.cloud.microsoft and Anthropic’s status page at about 10:55 a.m. Manila (UTC+8). The feed post does not repeat the incident ID; its text covers Exchange Online and the other Microsoft 365 services listed under MO1550821. Start times come from StatusStack’s copy of Microsoft’s Service Health records. Times are converted from UTC. We do not add a cause beyond what Microsoft has posted. The advice under What to do now is our own. No affiliate links. How we work.

Spotted an error? See our corrections policy and tell us.