ChatGPT, Claude and Grok broke down within hours of each other, and no link was confirmed
ChatGPT, Claude and Grok, three of the most widely used chatbots, went down within hours of each other on Thursday. OpenAI blamed a routing error, Anthropic said an infrastructure issue knocked out Claude.ai, Claude Code, Claude Cowork and its API for three hours, and SpaceX's Grok status page reported ongoing trouble. Cloudflare denied any fault of its own, and neither OpenAI, Anthropic nor xAI said whether the outages were connected.
Artificial Intelligence··Night
Three chatbots failed within a few hours of each other
The Verge reports that ChatGPT began returning errors to users at about 11.00 a.m. Eastern time on Thursday, with OpenAI's own status page describing elevated error rates across ChatGPT and Codex. The disruption reached beyond conversation to logins, file uploads, voice mode, search, deep research and image generation. Anthropic's chat interface, Claude Code and its API failed in the same window; the company's CJ Avilla attributed the partial outage to an infrastructure issue, and it was resolved by about 12.15 p.m. Eastern. Grok broke on Android, iOS and the web at 9.30 a.m. Eastern, telling people who prompted it on X that the model was overloaded.[1]
OpenAI, Anthropic and SpaceX gave The Register three separate explanations
The Register reports that OpenAI attributed the trouble to a routing error that began around 7.43 Pacific time and made ChatGPT and Codex unavailable for some users, with a fix in place by about 8.17 Pacific time, though the company's status page still showed elevated errors when the outlet filed its story. Anthropic told the outlet that an infrastructure issue took Claude.ai, Claude Code, Claude Cowork and the Claude API offline together, lasting three hours and six minutes before service was restored at 16.16 UTC, with elevated error rates hitting Sonnet 5 and other models. SpaceX's status page for Grok said it began investigating an outage at 6.30 Pacific time and was still working to restore service.[2]
No link was confirmed, and Cloudflare denied any role
Neither The Verge nor The Register obtained confirmation of whether the three failures were connected. The Register notes that Cloudflare, used in different capacities by all three companies, told the outlet its own systems were unaffected, and that AWS, Google Cloud and Microsoft Azure showed no matching problems on their status pages at the time. The Verge reports that OpenAI, Anthropic and xAI did not say whether the outages were connected, leaving the coincidence of three major chatbots failing within hours of each other unexplained.[1], [2]
