On the evening of March 2, 2026, Anthropic's Claude experienced a global service outage, affecting website login and many other functions. According to the official status page, the issue began appearing at 11:49 UTC and has since been resolved and is now under monitoring. The incident affected multiple regions, with numerous users reporting login failures and server errors, impacting users who rely on Claude for development, content generation, and workflows. Bloomberg reported that thousands of users were affected.
Event timeline: Approximately 4 hours from investigation to repair.
According to official information from status.claude.com , the events unfolded as follows (UTC):
- 11:49: System error begins; users experience login failures and error messages.
- 12:21: Official confirmation: The issue is related to the login/logout path of Claude.ai; API functionality is normal.
- 13:22: The engineering team has identified the problem and begun implementing the fix.
- 14:05: Some API methods are experiencing exceptions; ongoing processing.
- 14:42: Synchronous confirmation that Claude Opus 4.6 has encountered elevated errors.
- 15:25: Repair complete, entering monitoring phase.
The overall anomaly lasted approximately 3 to 4 hours, making it one of the larger service incidents since 2026.
The Claude outage affected the website version and Claude Code; the API remained relatively stable.
This anomaly primarily affects the Claude website and Claude Code, with some models (including Opus 4.6) experiencing an increased error rate. Many users have also reported login loops, HTTP 500/529 errors, connection interruptions, or request failures. However, the Claude API remains relatively stable, and the console platform.claude.com functions normally. The official statement emphasizes that this issue stems from a platform-level and validation process anomaly, rather than a failure of the models themselves.
Global user feedback: Workflow disruptions, developer impact
Numerous users on X (formerly Twitter) have reported being unable to log in, causing development work to stall, Claude Code to become unusable, and content and research tasks to be interrupted. Some users stated that Claude has become a core tool in their daily work, therefore the impact of the outage is far greater than that of a typical website malfunction.
The cause has not yet been announced, but speculation in the community suggests it could be due to an anomaly in the cloud infrastructure (Claude primarily runs on AWS), issues with a Middle Eastern data center, a surge in traffic, or an external attack.
This article is the latest! Claude suffered a global outage lasting over four hours, affecting thousands of users. Anthorpic claims the issue has been fixed. It first appeared on ABMedia ABMedia .
