Claude Experiences Brief Outage; Anthropic Still Monitoring Select Models

According to multiple tech media reports, Anthropic's AI chatbot Claude experienced a brief outage between March 2 and 3, 2026, affecting claude.ai web, mobile app, and Claude Code among other services. Users reported nearly 2,000 incidents on Downdetector, peaking Tuesday afternoon ET. Anthropic confirmed on its status page that it is investigating increased error rates across several models, API, Claude Code, Claude Cowork, and other components. The company attempted fixes and continued monitoring, but some models still exhibited anomalies.

The outage came shortly after Claude topped the App Store free app chart, driving a surge in user demand. Earlier, Anthropic drew public attention by refusing a Pentagon request to use AI for surveillance and weapons systems, prompting a wave of new users that strained servers. In an official statement, Anthropic attributed the issue to "unprecedented demand" and emphasized that enterprise API customers and businesses with integrated Claude models were unaffected. By March 3, services had largely recovered, though Anthropic said it would continue monitoring a few models to prevent further instability.

This incident highlights the stability challenges AI services face during high-traffic periods. As public interest in AI ethics debates grows, Anthropic's ability to balance technical openness with safety boundaries — and whether its infrastructure can keep pace with the traffic surge from its reputation breakout — will be a key test in the near term. The industry will also watch how the company optimizes load capacity to avoid similar disruptions affecting core developers and enterprise users.