Anthropic Outage Disrupts Claude API and Web Services

Anthropic took its core Claude services offline for approximately 36 minutes on August 16, 2026. The disruption began at 21:58 UTC with authentication failures and quickly escalated to degraded performance across five major products: claude.ai, Claude Console, the Claude API, Claude Code, and Claude Cowork. Users experienced sign-in failures, unresponsive interfaces, and stalled API requests. Anthropic deployed a fix and marked the incident fully resolved by 22:34 UTC, restoring all services, including Claude for Government.

This brief disruption marks the fourteenth incident logged on Anthropic’s status page since August 3, 2026. The preceding weeks saw a flurry of issues affecting various models and interfaces. For instance, Claude Fable 5 experienced elevated errors from August 14 to August 15. August 14 alone saw three incidents, including a 24-minute disruption to the Claude API, Claude Code, and Claude Cowork, alongside an invalid certificate issue on status.claude.com and overnight degradation of Claude Code Mobile, Routines, and Cowork Remote. Earlier, on August 13, Claude Mythos 5, Claude Fable 5, and Claude Sonnet 5 suffered errors for 95 minutes, while an August 12 incident degraded multiple models, primarily Fable 5, for over four hours. Other disruptions occurred on August 3, 4, and 5, involving OAuth failures and performance drops across Mythos 5, Fable 5, Opus 5, and Sonnet 5.

Despite this recent cluster of failures, Anthropic’s 90-day availability metrics remain above 99 percent for most services. Claude.ai sits at 99.34 percent uptime, Claude Code at 99.36 percent, Claude Cowork at 99.45 percent, and the Claude API at 99.44 percent. Claude Console maintains 99.82 percent, while Claude for Government stands at 100 percent. However, these fractions translate to significant real-world downtime. For developers, a 99.44 percent API uptime means roughly 12 hours of failed or degraded requests over a 90-day window. This reality underscores why multi-provider fallback routing has become an essential architecture for enterprise applications.