A Claude outage disrupted access to Anthropic’s AI assistant twice within a single day, leaving many users unable to sign in and triggering intermittent failures across key entry points. The first disruption was reported on March 2, with broad regional impact over a multi-hour window, followed by a second wave of availability issues on March 3.
Anthropic’s status updates indicated an ongoing investigation beginning at 03:15 UTC on March 3, with subsequent messages noting continued investigation through at least a 04:43 UTC update. During the interruption, users reported seeing recurring error prompts and “try again later” style messages while attempting to use the service.
What Was Reported as Affected
- Claude.ai and companion apps
- The developer console
- Claude Code
- Elevated errors across multiple Claude services, with the incident described as unresolved at the time of some reports
Outage-tracking snapshots cited thousands of user reports in the United States during the March 3 disruption and hundreds in India, with complaints concentrated around Claude chat, plus smaller shares tied to the app and website.
A notable reporting discrepancy remains: some accounts framed the disruption as impacting API-based usage, while another explicitly stated the API stayed up. That difference changes whether the primary impact falls on end users or API-dependent products.
