Claude AI goes down, disrupting global API and developer workflows

Anthropic’s Claude services are facing significant technical issues, causing widespread failures across the web platform, mobile and desktop apps, Claude Code, Claude Cowork, and its API. 

On September 29, 2026, users trying to access the platform were hit by loading errors, dropped requests, and unexpected logout prompts. The company acknowledged the disruption on its status page shortly after, confirming that API endpoints were impacted as well. While Anthropic has not released the cause of the downtime or an estimated time for recovery, it noted that repeating failed requests may occasionally work.

This latest system breakdown follows another major outage just days earlier on September 22, which disrupted several core models for over an hour and a half before engineers restored normal operations. 

However, this current incident spans a much broader share of Anthropic’s ecosystem, hitting both consumer chat features and enterprise tools that power external business applications.

The downtime temporarily halted workflows for individual users, researchers, and organisations reliant on Claude for content generation, coding assistance, and data analysis.

​As artificial intelligence becomes deeply embedded in daily operations, outages like this underscore the vulnerabilities of relying on single-provider infrastructure. Until Anthropic restores full system stability, affected teams must navigate these operational delays, which is a stark reminder of the need for reliable multi-model fallback strategies in mission-critical environments.