Claude Outage History

Users experienced delayed Claude.ai responses, with replies hanging for about 5 seconds after streaming before the message completes. The issue affected the Claude.ai component, while the Anthropic API was not affected. Impact began around 8:00 PM UTC.

Official updates documented the progression: at 11:19 AM UTC the Claude.ai component changed to a degraded state (warn) with the note 'Delayed response completion'; a follow-up at 12:39 PM UTC said a fix was being implemented; the service was restored by 1:13 PM UTC.

Overall, users experienced about 41 hours of degraded Claude.ai performance before restoration was completed by 1:13 PM UTC.

Impact began at 12:09 AM UTC when Claude Opus 4.6 started showing elevated error rates. The issue affected multiple Claude components, including Claude API (api.anthropic.com), claude.ai, platform.claude.com, and Claude Code, leading to failed requests and degraded performance for users relying on Claude Opus 4.6. The StatusGator updates showed ongoing investigation and efforts to identify the root cause, with the issue persisting into the early hours across Opus and Sonnet 4.6 and causing degraded experiences for users.

The Claude service briefly experienced an outage impacting API access (api.anthropic.com), Claude AI, and Claude Code. Users attempting to run tasks against Claude API or interact with Claude AI or Claude Code encountered errors and timeouts as elevated issues were reported across all three components. The incident began at 4:04 PM UTC with initial error reports and an investigation, and the services were restored by 4:18 PM UTC.

Users around the world experienced degraded Claude chat performance as responses slowed and timeouts occurred. The disruption affected Claude API (api.anthropic.com), claude.ai, and Claude Code, with impact beginning around 3:23 PM UTC. Observations from users described messages like 'Taking longer than usual' and noticeable delays during conversations and code tasks.

StatusGator recorded an Early Warning Signal at Mar 19, 2026 2:43 PM UTC, noting slow or timing out chat responses, preceding Claude's official acknowledgement. The status page then indicated elevated errors around 3:38 PM UTC and the services returned to normal by 3:48 PM UTC. The early-warning window closed by 4:03 PM UTC. The incident suggests a brief degradation lasting roughly the span from the onset to rapid recovery, with a notable burst of user reports in the afternoon before stabilization.

StatusGator detected issues starting around Mar 19, 2026 12:00 AM UTC with a surge of user reports by 12:15 AM UTC, signaling a widespread impact on Claude surfaces. The Claude API (api.anthropic.com), claude.ai, platform.claude.com, and Claude Code experienced authentication and chat failures, leaving users unable to sign in, access conversations, or send messages. Official updates show an escalation from a surface-wide warning to a full outage: the incident moved to a degraded/partial outage at Mar 19, 2026 12:28 AM UTC and then to a complete outage at Mar 19, 2026 12:33 AM UTC as engineers investigated. The disruption persisted for more than an hour, with recovery gradually progressing and services returning to normal by 1:43 AM UTC. StatusGator’s early signal highlighted the issue approximately 12:00 AM UTC with a lead time of about 33 minutes before the official acknowledgment.

An outage affected Claude.ai and Claude Code, causing elevated errors and failed prompts across users worldwide. API calls returned errors and login/logout actions were impacted, making Claude Code less usable and causing prompts to not respond in many regions. StatusGator detected the issue at Mar 18, 2026 3:14 PM UTC and the Early Warning Signal ran until Mar 18, 2026 5:09 PM UTC, with the provider first acknowledging the disruption at 3:18 PM UTC. A fix was deployed by 3:28 PM UTC and recovery began, with Claude.ai and Claude Code returning to normal by 4:08 PM UTC. This disruption drew a surge of user reports around mid-afternoon, underscoring the widespread impact.

Users worldwide faced outages as Claude Opus 4.6 overloaded the platform, triggering widespread errors and timeouts across Claude API (api.anthropic.com), claude.ai, platform.claude.com, Claude for Government, and Claude Code. The incident began around 12:30 PM UTC and persisted into the afternoon, with a high volume of user reports from multiple regions indicating global disruption. StatusGator recorded an Early Warning Signal for overload starting at Mar 18, 2026 11:32 AM UTC and ending at Mar 18, 2026 1:49 PM UTC, including lead time of about Mar 18, 2026 12:33 PM UTC before the official acknowledgment.

An outage affecting Claude Opus 4.6 began around 6:43 AM UTC with elevated errors across Claude API (api.anthropic.com), claude.ai, platform.claude.com and Claude Code. Users across multiple regions reported slow responses, failed requests, and UI loading issues, indicating a broad impact on both programmatic access and the user interface. The disruption affected a global user base, with reports of degraded performance appearing for several hours.

The Claude Opus 4.6 outage began with widespread errors across Claude API (api.anthropic.com), claude.ai, platform.claude.com, and Claude Code. StatusGator users saw elevated failures in the Claude app and API starting around 7:37 PM UTC, with official acknowledgment around 7:48 PM UTC. The incident quickly escalated into a broader outage as users reported API errors, timeouts, and degraded performance. StatusGator’s Early Warning Signal indicated the issue ahead of the provider’s notice, showing a lead time of about 11 minutes between detection and official acknowledgment; the signal remained active from Mar 17, 2026 7:37 PM UTC until Mar 17, 2026 9:54 PM UTC. A large volume of user reports in the first hours underscored the global impact across multiple regions.

Mitigations were deployed and monitoring began around 10:43 PM UTC as engineers worked to restore services. A fix was implemented and results monitored, with all core components returning to normal by 12:28 AM UTC across Claude API, claude.ai, platform.claude.com, and Claude Code, marking the end of the incident. The outage disrupted workflows for developers and users relying on Claude across multiple fronts, including API access, UI interactions, and code tooling, during the evening and into the late night.

Users worldwide faced unresponsive Claude prompts as the chat service experienced a widespread outage. Reports spanned multiple regions including Ashburn, Frankfurt, Paris, Kuala Lumpur, Berlin, and Santiago, with chats taking longer than usual or failing to return any output. The outage began at 5:09 PM UTC and was resolved by 6:31 PM UTC. StatusGator detected the issue at Mar 17, 2026 5:09 PM UTC, ahead of any provider acknowledgement.

Claude's official status page did not show updates during the incident, leaving users to gauge impact from reports themselves. The surge in reports across continents confirms a global impact on Claude chat, with the core chat service degraded for about 1 hour and 22 minutes before gradual recovery.