Darwin
AI & Technology

Claude Login Outage Hits Claude.ai and Claude Code as Microsoft 365 Connector Fails

Anthropic reported login failures on Claude.ai and Claude Code on 13 November, while a separate Microsoft 365 outage broke Claude's Outlook connector.

Written byDarwin Corp
Published13 November 2026
Read3 min

Summary

Anthropic reported an authentication problem on Claude.ai and Claude Code from 09:58 UTC on 13 November, deployed a fix at 11:05 UTC and later marked the incident resolved. Separately, Claude's Microsoft 365 connector and Claude for Outlook were failing because of Microsoft's outage, which began on 12 November. Anthropic has not explained the login failure, and the Microsoft connector problem was still listed as open in the latest reports.

Anthropic opened an incident at 09:58 UTC on Friday 13 November after users had trouble authenticating to Claude.ai and Claude Code. A fix was deployed at 11:05 UTC, and Anthropic later marked the incident as resolved.

Anthropic has not explained what caused the login failure. Its notices also do not establish that Microsoft caused it. A separate incident affecting Claude's Microsoft 365 connector is caused by Microsoft's own outage and was still listed as open in the latest reports. This is a developing story based on reports from 13 November.

What happened

Anthropic's status page said it was investigating a problem with users authenticating to claude.ai and Claude Code. The incident record also listed the developer console among the affected products. The status page showed Claude.ai in a major outage during the incident, according to Startup Fortune. Anthropic did not give a technical explanation for the failure in the notices reviewed, and its API was listed as operational when checked.

Users also reported the problems on social media. One widely shared post described Claude Code logging users out mid-session and "OAuth request failed" errors. The scale of the disruption has not been independently measured.

The Microsoft 365 connector problem

Anthropic opened a separate incident, "Elevated errors affecting the Microsoft 365 connector", at 17:52 UTC on 12 November. Its 22:23 UTC update said the issue was caused by a Microsoft 365 service incident, and that mail, calendar and search tools in the connector were still failing for most requests. Claude for Outlook was also affected.

Users could therefore reach Claude itself but still fail to use Outlook mail, calendar and SharePoint features through it. The connector incident remained under a separate warning after the login incident closed.

Microsoft's outage

Microsoft acknowledged the problem at 12:54 pm ET on Thursday 12 November, saying it was investigating users' trouble accessing Exchange Online mailboxes in the Americas. It said the wider Microsoft 365 incident began at 1:26 pm ET and affected Exchange Online, Teams, SharePoint Online, Microsoft 365 Copilot and Defender XDR. Outlook error reports on Downdetector peaked at 128,874.

Reuters reported that the outage also hit the connectors that OpenAI's and Anthropic's AI services use to reach Microsoft accounts.

Sources disagree on whether the outage has ended. gHacks reported that at 12:00 UTC on 13 November Microsoft still said users were unable to connect to multiple services. A secondary write-up citing Astra Security said it was largely resolved around 05:33 UTC that day. Microsoft's own admin-centre notices (MO1550821 and EX1550543) hold the latest status.

What to watch

  • Whether Anthropic publishes a cause for the login incident.
  • Whether the Microsoft 365 connector and Claude for Outlook recover once Microsoft's service stabilises.
  • Updates on status.claude.com and Microsoft's service health notices.

Analysis

This section is analysis, not reporting.

The two events were separate, though they overlapped, and the distinction matters for users. One affected sign-in to Claude itself, and the other affected a third-party connection that Claude depends on. Teams that rely on a single AI provider, or on one login route, have a single point of failure in either case. Anthropic's lack of a stated cause means the login problem's origin is not yet known.

Sources

Frequently asked questions

Anthropic opened an incident at 09:58 UTC for problems authenticating to claude.ai and Claude Code. Anthropic later marked the login incident as resolved.

Not according to Anthropic's notices. The Microsoft 365 connector problem is attributed to a Microsoft service issue, but Anthropic's notices do not establish that Microsoft caused the later authentication problem.

Anthropic opened a connector incident at 17:52 UTC on 12 November. Its latest notice said most attempts to use the connector's email, scheduling and search functions were failing, and Claude for Outlook was also affected.

Anthropic marked the login incident resolved, and its dashboard listed the API as operational when checked. The Microsoft 365 connector remained under a separate warning. Check status.claude.com for the latest.

Outlook error reports peaked at 128,874 on Downdetector, with Exchange Online, Teams, SharePoint, Copilot 365 and Defender XDR all affected. Sources disagree on when the outage ended.