On March 2, 2026, Anthropic’s AI chatbot, Claude, experienced significant disruptions, affecting thousands of users globally. From around 11:49 UTC (6:49 a.m. ET), many reported difficulties accessing the service, including login failures and error messages across various platforms, such as claude.ai and Claude Code.
The company acknowledged the issue on its official status page, indicating that it was “currently investigating” elevated error rates. Subsequent updates revealed that the problems were primarily related to consumer-facing platforms, particularly the Claude.ai website and related login functions. Notably, the core Claude API remained operational, allowing some businesses to continue functioning without interruption.
By early afternoon UTC, Anthropic had identified the root cause and was actively working on a fix, although it did not provide an estimated time for full service restoration. Throughout the day, updates from Anthropic confirmed ongoing investigations, clarifying that while some API methods were initially affected, the backend models and API endpoints were largely unaffected.
According to Downdetector, a site that tracks user-reported issues, complaints peaked at nearly 2,000 reports around 6:40 a.m. ET. While complaints decreased to about one-third of that level by mid-morning, they remained significantly higher than normal. Users most frequently reported problems with the chat interface (approximately 39% of complaints), the mobile app (35%), and the website (15-20%), citing HTTP 500 and 529 errors, timeouts, and generic error messages.
Anthropic characterized the incident as a “partial outage,” emphasizing that it primarily affected web-based access and authentication, rather than the underlying AI models like Claude Opus 4.6 or enterprise integrations. Businesses using Claude through API or embedded systems appeared to be largely unaffected, allowing them to maintain operations.
Despite the company’s efforts, the precise technical trigger for the outage remained unclear. Anthropic refrained from speculating on potential causes such as server overload or configuration errors. Some users speculated that the disruptions could be attributed to “unprecedented demand” for Claude, especially as the chatbot had gained prominence amid competition with ChatGPT and improved app store rankings.
Social media platforms buzzed with user frustration, featuring memes and comments reflecting on the irony of relying on Claude Code to troubleshoot the very outages affecting Claude. The timing of the outage also coincided with increased scrutiny of Anthropic, particularly due to its recent tensions with U.S. government contracts. Reports indicated a termination of a contract with the Department of Defense over “safety alignment conflicts” and restrictions on military applications.
While some users drew speculative links between the outage and geopolitical events, such as alleged attacks on AWS infrastructure in the Middle East—Anthropic’s primary cloud provider—the company attributed the issues to internal web and login paths rather than broader cloud failures.
As of March 2, 2026, in the evening KST, partial recovery appeared underway for some users, though intermittent issues persisted with login flows and web access. Anthropic advised affected users to monitor the status page for real-time updates and recommended trying again later or using API routes where applicable.
This incident highlights the ongoing challenges in AI infrastructure. As technology advances, issues of front-end reliability, authentication systems, and traffic management remain critical, particularly during peak adoption phases. Similar outages have previously impacted competitors like OpenAI, often linked to surges in user demand or scaling challenges.
Users worldwide expressed frustration over workflow interruptions, especially those relying on Claude for coding, research, or daily productivity. With no confirmed timeline for complete service restoration, many turned to alternatives such as Grok and ChatGPT while awaiting fixes. Anthropic’s team continued to work on resolving the issues, hoping for a swift return to normal service levels. Historically, the company has maintained high uptime, making this event notable for its scope and visibility.
