Anthropic's Claude AI services experience global outage
Anthropic's Claude AI services experienced a global outage on 5 August 2026, affecting multiple model versions and disrupting access for users, developers, and corporate systems. The company identified elevated error rates and worked to restore service across its platforms.
What happened
Claude, Anthropic's AI assistant, experienced global instability on 5 August 2026, affecting the web and mobile versions of the platform and API connections used by developers and corporate systems.
The company identified elevated error rates in requests made to Claude Mythos 5, Claude Fable 5, Claude Opus 5, and Claude Sonnet 5 models. Anthropic's technical teams worked to apply repairs to the system.
Reported impact
Users reported difficulties sending prompts, receiving responses, and using resources linked to the platform, including Claude Code, a tool for programming. Downdetector recorded over 127 notifications of problems in Brazil at the start of the morning, with alert volume also showing upward trend in the United States.
Governance controls
Anthropic confirmed the instability on its official Claude Status page and classified the incident as degraded performance in multiple models. The company stated it had identified the origin of the problem and provided updates as the service was restored across affected platforms.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.