Claude AI experiences global outage, users report service issues

Key Takeaways

• Anthropic’s Claude AI assistant experienced a widespread global outage affecting its web and code-related services.
• The disruption, which began around 2:10 AM ET, led to slow responses, instability, and complete service unavailability for many users.
• Both free and paid users across various interfaces, including web chat, mobile access, and Claude Code, were impacted.
• Anthropic reported deploying a fix, but recovery was gradual, with some users continuing to experience issues even after services were declared operational.
• The incident underscores the importance of multi-vendor strategies for organizations reliant on AI services.

Anthropic’s Claude AI Suffers Global Service Disruption

Anthropic’s prominent artificial intelligence assistant, Claude, faced a significant worldwide service interruption today, leaving users reporting widespread issues with its web and code-centric functionalities. Reports indicated that services were either sluggish, unstable, or entirely inaccessible for a considerable period.

The disruption commenced approximately at 2:10 AM ET (7:10 AM GMT), marked by a sharp increase in incident reports on Downdetector across numerous geographical regions. At its peak, the platform recorded over 200 user reports in the United Kingdom and around 140 in the United States, signifying a broad impact on both complimentary and subscription-based users.

Service Issues and Remediation Efforts

Initially, Anthropic’s official status page categorized the event as a “partial outage,” acknowledging heightened error rates across critical Claude services. The company confirmed that a resolution was actively being implemented. Subsequent updates to the status page indicated that a remediation had been deployed and was under active monitoring. Despite these assurances, a significant number of users continued to encounter failed or delayed responses, suggesting a phased recovery rather than an immediate return to full functionality.

The outage’s scope was not confined to a single access point, with users detailing problems across Claude’s web interface, mobile applications, and Claude Code environments. Anthropic later specifically identified “elevated errors on Opus 4.6,” its flagship model launched in February. Users also reported that sessions utilizing Sonnet 4.6 would indefinitely hang, displaying messages such as “gathering my thoughts” or “still working on it.” While some users eventually received responses after considerable delays, others experienced outright request failures or timeouts.

A contentious aspect of the outage emerged for users of the free tier, where some received a message stating, “due to unexpected capacity constraints, Claude is unable to respond to your message,” followed by a prompt to upgrade to a Pro plan. This specific behavior was widely criticized as ill-timed, given that the underlying problem appeared to be platform instability rather than typical capacity limitations.

Recovery and Industry Implications

Over time, Downdetector’s graphs began to reflect an improvement, with reports from the UK decreasing from hundreds to a few dozen, and US reports also declining as mitigation strategies took effect. Anthropic’s status page eventually declared “all systems operational.” However, some users still observed slower-than-normal response times and intermittent failures shortly after this update.

From the perspective of reliability and cybersecurity operations, this incident underscores an escalating risk associated with over-reliance on a singular AI provider for essential workflows. As organizations increasingly transition between AI vendors, such as the recent shift observed from ChatGPT to Claude, outages like this unequivocally demonstrate the critical need for diversified multi-vendor strategies, robust fallback mechanisms, and transparent incident communication when core AI infrastructure experiences disruptions.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.