Microsoft Outage Hits Outlook and Teams, Reports Near 140,000

Outage reports for Microsoft's Outlook peaked globally at nearly 140,000 around 18:00 GMT on Thursday, according to monitoring website Downdetector, as the company fought to restore services including Teams, Outlook and its AI assistant Copilot.

Microsoft blamed "connectivity problems" for the failures. "We're aggressively pursuing all pathways to reduce impact," the company wrote on X. The scale matters: for context, 140,000 individual reports typically represents a far larger base of affected users, since most people never file a report at all.

The trouble started in the US, but Microsoft said users "in any region" may be affected. That wording signals a systemic infrastructure problem, not a regional blip.

Why did the recovery keep slipping?

Microsoft's own updates explain the difficulty. The company said part of its infrastructure was "experiencing elevated load across multiple regions," which was "likely contributing to the connectivity errors."

A later update sharpened the diagnosis: the extra load was probably caused by several servers being "unexpectedly unavailable; straining the remaining infrastructure."

Microsoft managed to restore part of its service. But about an hour later it warned that those servers "can become unresponsive again once they receive traffic." In plain terms, the fix wasn't stable — every time users returned, the recovered systems risked buckling under renewed demand. It's the classic death-spiral pattern: capacity comes back, traffic floods in, capacity collapses again.

What does this change for AI users?

The outage didn't stop at Redmond's own products. It rippled into two of Microsoft's highest-profile AI partners.

OpenAI said it was investigating an issue with Microsoft connectors in which new connections may be broken and tools are not working as expected. Anthropic said it had identified an issue with an upstream provider affecting its Microsoft 365 connector.

For practitioners, that's the practical takeaway: if your AI workflow plugs into Microsoft 365 through either vendor's connector, this outage reached you even if your own Teams or Outlook session looked fine.

A pattern, not a one-off

This is the second significant Microsoft infrastructure event in quick succession. Last month, the company reported an outage of its Azure cloud platform that impacted its suite of productivity software and a range of industries worldwide.

The real-world cost is already showing up in corporate calendars. Electric truck maker Workhorse Group, which was expected to report third-quarter results on 12 November, rescheduled the release to 16 November, citing business disruptions caused by the global Microsoft outage. When a public company moves an earnings date four days because of a software vendor's outage, that's a measurable business impact — and a data point boards will notice.

What happens next?

Watch two things. First, whether Microsoft's restored servers hold under full traffic — the company's own warning about servers becoming unresponsive again suggests fragility until load stabilizes. Second, whether the Azure outage and this event share a root cause; two multi-region incidents in as many months will put pressure on Microsoft to explain what's changed in its infrastructure posture.

For companies running operations on Teams, Outlook and Copilot, the episode is a reminder that even partial dependency on a single vendor's cloud creates single points of failure. Expect renewed conversations about continuity planning in the weeks ahead.