Anthropic's Claude Hit by Global Outage

The March 2nd outage primarily impacted consumer-facing services, including the main claude.ai web interface, Claude Code, and user authentication paths, while the core API remained largely operational. Users reported encountering HTTP 500 and 529 errors, login failures, and timeouts starting around 11:49 UTC, with thousands of reports logged on monitoring sites. While Anthropic confirmed it was investigating "elevated errors," it has not publicly disclosed a specific technical root cause for the disruption. This incident occurred just as Claude's iOS app surged to become the most downloaded free app in the U.S., overtaking ChatGPT on March 1st. The rapid climb from outside the top 100 in late January was reportedly fueled by public backlash against OpenAI following its announcement of a partnership with the U.S. Department of Defense, a deal Anthropic had publicly refused over ethical concerns regarding mass surveillance and autonomous weapons. Anthropic's infrastructure is built on a multi-cloud strategy, a key differentiator from competitors who often standardize on a single provider. While Amazon Web Services (AWS) serves as its primary cloud partner for model training, the company has also made massive investments in Google Cloud to leverage its specialized Tensor Processing Units (TPUs). A November 2025 deal with Microsoft and Nvidia adds Azure to the mix, making Claude the only major model available across all three top cloud platforms. This multi-cloud and multi-accelerator (Nvidia GPUs, AWS Trainium, Google TPUs) approach provides significant operational flexibility and is designed to mitigate risks associated with reliance on a single architecture. However, the platform is not immune to stability issues. Records show a history of smaller incidents in February 2026, including elevated error rates on specific models like Opus 4.6, and a more significant multi-day outage that affected multiple models and services. The outage highlights the reliability stakes in the increasingly competitive enterprise AI market. For enterprise search, startups like Glean and Hebbia are key competitors. Glean focuses on unified search across 100+ workplace apps, using a knowledge graph to personalize results based on organizational structure. Hebbia, by contrast, is purpose-built for deep analysis of unstructured documents in knowledge-intensive sectors like finance and law. Both Glean and Hebbia utilize Retrieval-Augmented Generation (RAG) but apply it differently. Glean builds a broader enterprise knowledge graph, while Hebbia focuses on deep document analysis and synthesis. Cohere, another major player, provides its language models for businesses to build their own applications, offering flexible deployment across various cloud environments or on-premises.