Overview
- Anthropic logged four discrete incidents starting at 19:49 BJT on March 2 and reported full recovery by about 05:16 BJT on March 3.
- Consumer access to claude.ai, mobile apps, and some API routes saw failed requests, timeouts, and unstable results as the company temporarily took these services offline.
- Anthropic said enterprise customers with internal integrations were not impacted during the disruptions.
- Downdetector recorded roughly 1,952 U.S. problem reports at the peak, and Anthropic said it is investigating causes while working on capacity.
- The outages followed sharp growth in usage, with company data showing free users up over 60% since January and paid subscribers doubling since October, alongside third‑party reports of rising Claude downloads and a spike in ChatGPT app uninstalls.