Claude disruption: Anthropic services down for about 40 minutes

Anthropic confirmed a major disruption of its Claude services on Saturday evening. From 21:58 UTC, the status page reported authentication issues affecting Claude.ai, Claude Code and Claude Cowork; at 22:07 UTC it added "degraded performance" on Claude.ai and platform.claude.com. After roughly 40 minutes, operations were reportedly largely back to normal.

Anthropic initially described the incident tersely: it was investigating "an issue preventing some users from authenticating to Claude.ai, Claude Code, and Claude Cowork." Affected users reported failed logins, interfaces failing to load, and interrupted requests.

Important for businesses: the Claude API and Console remained operational according to Anthropic – production integrations via the API were not hit to the same degree as the consumer-facing products.

The company did not initially name a root cause; both incidents remained "under investigation" at the time of reporting, per BleepingComputer. The outage joins a series of brief disruptions at major AI providers this year.

The incident underlines how central AI assistants have become to daily workflows: even a sub-one-hour disruption triggered outage reports worldwide and renewed discussion about failover strategies.