Is Claude Down? Anthropic says claude down outage restored after hours of elevated errors
Claude, Anthropic’s AI chatbot, experienced an outage on Monday that left consumer-facing services unavailable and prompted a company investigation; the wording "claude down" appeared widely as users reported connection problems. Anthropic later said the service was restored and that teams were working to match unusually high demand tied to recent surges and a public government dispute.
Is Claude Down? Timeline from early morning reports to restoration
Users began reporting widespread problems shortly after 6 a. m. ET, and Anthropic’s status page acknowledged an issue at 11: 49 a. m. GMT (6: 49 a. m. ET) saying it was "currently investigating this issue. " An update at 9: 49 a. m. ET indicated the company had identified the issue and was implementing a fix, and around 11 a. m. ET Anthropic said service had been fully restored after more than four hours of technical issues. The company later issued an update on Mar. 2, 2026, at 4: 37 p. m. UTC saying Claude should be running as usual while the team continued to match incredible demand.
Which systems failed and which remained functional
Anthropic said the outage produced "elevated errors" that affected the Claude chatbot and several named components: Claude Opus 4. 6 (the latest-generation large language model), Claude Console (the developer platform), and Claude Code (the coding assistant). Company posts noted problems with the login and logout paths and that the team had discovered some API methods were not working. At the same time, Anthropic said the Claude API that powers business customers remained unaffected.
User impact: reports, signups and download rankings
More than 2, 000 users reported issues to an outage tracker during the disruption, while Anthropic said daily signups were breaking records and free users had risen more than 60% since January. The outage came hours after Claude reached the top of the U. S. free-app download chart, with competitors ChatGPT in second place and Google’s Gemini in fourth.
Government split and military technology shifts after the blowup
The extended downtime followed a public clash between Anthropic and U. S. officials: President Donald Trump instructed U. S. government agencies to stop using Anthropic’s services, citing disagreements with the Pentagon over how the model could be used for wartime actions and writing, "The United States of America will never allow a radical left, woke company to dictate how our great military fights and wins wars!" An agreement announced later gave the U. S. government six months to phase out the software and instead adopt OpenAI's GPT models; OpenAI also entered into a deal to provide its technology for military use in "classified environments. " Despite the dispute, Anthropic’s Claude was used during U. S. strikes on Iran early Saturday.
Security concerns, industry reaction and ongoing investigations
Separately, employees at other major AI companies published an open letter standing by Anthropic, and it has seen unprecedented demand in recent days. Security fallout has also surfaced: a hacker used Claude to attack Mexican government agencies and stole tax and voter data, and Anthropic identified parties it called DeepSeek, Moonshot and MiniMax as responsible for distillation attacks. Meanwhile, reports indicated the U. S. Department of Defense reached a deal to use Grok in classified systems, and market watchers were focused on a major chipmaker’s fourth-quarter results amid continued AI concerns.
What Anthropic told users as service returned
Anthropic expressed gratitude to users and said the team was working to restore full service while matching high demand. The company stated that consumer-facing surfaces were temporarily unavailable and that it would share updates as work continued. The last comparable incident of "elevated errors" had occurred on Saturday, Feb. 28, 2026, with problems then found and resolved affecting Claude Opus 4. 6.
