The Claude AI outage on April 15 disrupted access for thousands of users on Wednesday, with reports of issues across Claude Code, Claude chat, the mobile app, and user logins. Outage reports rose sharply during the morning, while Anthropic moved from investigating the issue to rolling out a fix.
More than 15,000 users reported problems, with connectivity issues described as the main complaint. The disruption began in the early morning hours, and by 8:02 a.m. PT, more than 7,000 reports had already been logged, after which the total climbed rapidly over the next 25 minutes.
The breakdown of reported issues showed the outage affecting multiple parts of the service. According to the provided report, 46% of users flagged problems with Claude Code, 24% cited issues with Claude chat, and 18% reported problems with the mobile app. Users also reported login failures and issues accessing the website.
Downdetector-based reporting also showed thousands of user reports as the incident unfolded.
Anthropic’s status updates indicated the company first said it was investigating the issue, then later changed the incident status to say a fix was being rolled out. A mirrored status feed also reflected that the issue had been identified and that recovery work was underway.
By 8:25 a.m. PT, Anthropic said the Claude API had fully recovered as of 8:01 PT / 16:01 UTC. However, the same update indicated that problems with Claude AI logins were still being mitigated, while logged-in Claude Code users could still use the service.
The latest updates suggest the outage was at least partly stabilised, especially for the API. Still, because login and access issues were part of the disruption, users may continue monitoring official service updates for confirmation that all affected tools have fully returned to normal
