On November 13, 2026, users of the Claude AI platform experienced a massive outage affecting key services from Anthropic. According to the developer, the issues impacted authentication in the Claude.ai web interface and the Claude Code tool, as well as the operation of the developer platform platform.claude.com. The incident quickly drew community attention because Claude is actively used in both corporate and personal workflows, and its unavailability directly impacts users' daily routines.
What exactly stopped working
According to reports from Anthropic, the failure was not limited to a single component: users could not log in to Claude.ai accounts, start sessions in Claude Code, or access the API via platform.claude.com. This effectively disabled the "user-model" link for a significant portion of the audience. Separately, the company noted that tools for email, calendar, and search continued to work with errors via the Claude connector, and the Claude for Outlook add-in would not open at all for some users or failed to perform actions on emails and calendar events.
Cause: Failure on the Microsoft 365 side
In a separate clarification, Anthropic explicitly stated that the integration failure with Microsoft 365 was caused by a problem on Microsoft's end. This means the root cause was not in Claude's infrastructure, but in the dependency on external services within the Microsoft product suite. The company directed users to the official Microsoft 365 status page, where recovery updates are published, which allowed for real-time tracking of the incident and ensured it was not confused with internal AI platform failures.
Recovery and incident status
Upon completion of the work, Anthropic confirmed that the problem was resolved and the incident was given the status "resolved." This means authentication, Claude Code, the platform.claude.com site, and integration scenarios with Microsoft 365 have returned to normal operation. The company recommended that users check the Microsoft 365 status page if necessary to confirm the full recovery of third-party services on which the integrations depend.
Context: Vulnerability to external dependencies
The November 13 incident clearly demonstrated how modern AI products depend on the ecosystem of external services, even when the main computational engine is functioning correctly. Integrations with corporate tools — email, calendar, search — become a critical point of failure if their provider experiences an outage. For Anthropic, this is an additional argument for transparent communication: the company not only identified the failure but also publicly pointed to the source of the problem, directing users to the authoritative Microsoft 365 status source.
