Updated ETGet the DI Daily by Email

Technology news for e-commerce professionals, with our take on why it matters.

Microsoft 365 outage disrupts Outlook and Teams, and ChatGPT and Claude connectors

Microsoft 365 outage disrupts Outlook and Teams, and ChatGPT and Claude connectors
Image: Marketing4eCommerce México

Downdetector logged more than 130,000 Outlook reports at the peak, and Microsoft tied the failure to servers failing health checks, Marketing4eCommerce México reported.

Source: Marketing4eCommerce México

A Microsoft 365 outage on Thursday, November 12, 2026, disrupted access to Outlook, Teams, OneDrive, SharePoint Online and Microsoft Copilot, Marketing4eCommerce México reported. Early reports were concentrated in North America, and Microsoft confirmed the incident could affect connectivity in other regions. Downdetector logged more than 130,000 Outlook reports at the peak, about 46,000 for Microsoft 365 and thousands for Teams, with the total across platforms above 170,000. The outlet noted those figures count user notifications, not necessarily affected individuals.

According to Microsoft's updates as relayed by the outlet, the problem was not initially tied to unusual traffic but to several servers that failed automated health checks and were pulled from service, which overloaded the remaining infrastructure. Microsoft paused deployments and configuration changes, redirected traffic and restarted servers. It had not published a definitive cause or a time for full restoration when the report was last updated.

The outage also reached AI tools. OpenAI said it was investigating errors in its Microsoft connectors that could block new enterprise connections. Anthropic said an upstream provider issue was affecting its Microsoft 365 connector, with its status page showing errors on most requests for email, calendar and search tools, and some Claude for Outlook functions affected. ChatGPT and Claude themselves kept running, the outlet reported.

Read the full article at Marketing4eCommerce México

Our take

AI assistants connected to company email and files inherit those systems' outages, so a Microsoft failure can stall support and operations work running through ChatGPT or Claude.

  • Map which AI workflows depend on Microsoft 365 or another single provider.
  • Keep a manual fallback for customer email and support during outages.
  • Watch for Microsoft's root-cause explanation, which had not been published.

Was our take useful?

Key takeaways

  1. Downdetector logged more than 130,000 Outlook reports at the outage's peak
  2. OpenAI and Anthropic both reported failures in their Microsoft 365 connectors
  3. Microsoft linked the outage to servers failing health checks, per the report