On June 23, 2026, Anthropic’s Claude AI platform experienced a significant service disruption, leading to elevated error rates across several of its models. The issue was first acknowledged at 14:19 UTC, with the company initiating an investigation into the cause. By 14:25 UTC, Anthropic identified the problem and began implementing a fix. By 14:53 UTC, the fix was applied, and the company entered a monitoring phase to ensure system stability.
The outage impacted multiple services, including the consumer-facing claude.ai interface, the Claude Console at platform.claude.com, the Claude API, Claude Code, and Claude Cowork. Notably, Claude for Government remained operational throughout the incident. Users reported partial outages, indicating that while some requests failed or returned errors, others were successfully processed.
This incident is part of a series of disruptions affecting Anthropic’s services in June 2026. Earlier on the same day, another issue involving elevated errors for Claude Opus 4.8 occurred between 06:28 UTC and 08:45 UTC. Additionally, on June 22, multiple incidents were logged, affecting models such as Opus 4.8, Opus 4.7, Opus 4.6, Sonnet 4.6, and Haiku 4.5. On June 16, a two-phase incident resulted in error rates reaching approximately 10% across all Sonnet and Opus models during part of the disruption.
These recurring outages highlight the challenges in maintaining the reliability of AI services as they become increasingly integral to various applications. Organizations relying on AI platforms like Claude must consider implementing robust contingency plans to mitigate the impact of such disruptions on their operations.
