• Sources: primary, discussion, discussion, discussion
  • Summary: Anthropic began investigating failed authentication to claude.ai and Claude Code at 09:58 UTC on 2026-11-13. A fix was deployed at 11:05 UTC and the incident was marked resolved at 11:58 UTC. The status page lists claude.ai, Claude Console, Claude Code and Claude Cowork as affected.
  • Why it matters: Teams that run Claude Code with account sign-in had no fallback during the window, which argues for API-key or second-provider failover in agent workflows.
  • Background: Claude Code can authenticate either through a claude.ai account login, which uses OAuth, or through an API key from the Console. An authentication fault on the account side blocks interactive sessions even when model serving is healthy. The incident came while a separate Microsoft 365 outage was still unresolved.
  • Comments: The main Hacker News thread drew 292 points and 243 comments, with a second Claude Is Down thread at 146 points and 125 comments and a Reddit discussion hub. Commenters who use only the API said they switch to another model provider in seconds, while subscription users described pre-tuned failover across plans from several labs.