Microsoft 365 outage hit nearly 140,000 Outlook users worldwide

Outage reports for Outlook alone climbed to almost 140,000 at around 18:00 GMT on Thursday, as a widespread Microsoft 365 failure cut across Outlook, Teams, OneDrive and the Copilot AI assistant. Users in the US, Japan and Singapore were among those reporting problems on the outage tracker Downdetector.

Microsoft traced the disruption to a group of servers that unexpectedly went offline. The failure pushed extra load onto the remaining infrastructure until that infrastructure also began to struggle. The company's explanation points to a capacity and resilience problem rather than a security breach, and Microsoft has given no indication of a cyberattack.

What it has not done is say why those servers failed in the first place. That is the answer IT departments will be waiting for.

What did the outage actually break?

The impact varied from service to service. Microsoft says messaging and online meetings mostly kept working. Some users could not open files, saw calendar entries with missing details, or found Copilot unreliable.

Recovery was uneven. Microsoft posted updates on X saying it was making progress, but warned that restored services could stop responding again and that customers in any region might still be hit.

The effects reached beyond Microsoft's direct customers. OpenAI and Anthropic, both of which work with Microsoft's systems, were also affected. OpenAI said it was investigating problems with its connections to Microsoft services that had left some of its tools not working as expected.

The business cost was concrete. Electric truck maker Workhorse Group moved its third-quarter earnings report from 12 November to 16 November and blamed the outage.

What does this change for enterprise buyers?

This is not a one-off stumble. Just last month, an outage on Microsoft's Azure cloud platform took down its productivity suite and disrupted companies around the world. Two major incidents within weeks raise the question of how much risk organisations take on when email, chat, file storage and, increasingly, AI tools all depend on one provider. When that provider goes down, many teams have nothing to fall back on.

The stakes rise as Microsoft builds Copilot more deeply into everyday work. An outage now means employees lose an assistant they are being encouraged to rely on, not just their inbox.

Customers will want a detailed post-incident report covering three points:

• what failed
• why the backup capacity could not absorb the load
• what will change

Whether Microsoft provides that level of transparency may shape how hard enterprise buyers push on resilience at their next contract renewal. For practitioners, the practical takeaway is simple: two incidents in a month make fallback arrangements for email, chat and AI tooling a procurement conversation, not just an IT one.