Anthropic, OpenAI, and xAI experienced rare service outages Thursday morning, temporarily disrupting access to their AI chatbots and related platforms. SpaceX, xAI’s parent company, said Thursday afternoon that the Grok outage was caused by “an outage at the Memphis Compute Center this morning.”
Although the simultaneous AI outages initially appeared to be connected—potentially through shared cloud or infrastructure providers—neither OpenAI nor Anthropic identified an external cause in comments to WIRED. SpaceX did not respond to WIRED’s request for comment, but apologized publicly to affected computing partners. Anthropic and xAI announced a computing partnership with SpaceX in May.
OpenAI spokesperson Kathleen Chaykowski told WIRED that a routing error beginning at approximately 7:43 a.m. PT on Thursday, September 3, caused ChatGPT and Codex to become unavailable for some users across platforms. OpenAI said the issue was resolved by approximately 8:17 a.m. PT and that the fix continued to be monitored.
Anthropic declined to comment on the incident. The company reported a partial outage beginning at 6:23 a.m. PT, with increased errors affecting requests to Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5. Anthropic later said it had identified the cause and implemented a fix. The company marked the outage as resolved at 9:16 a.m. Pacific Time. Claude Sonnet 5 also appeared to experience a brief disruption shortly after 9 a.m. Pacific Time.
xAI reported that the Grok outage affected all platforms and services beginning at 6:30 a.m. Pacific Time. The company said it was investigating the problem on its service status page, which stated that “Grok is experiencing an issue and we are working to restore service as soon as possible.” At 10:05 a.m. Pacific Time, xAI marked the incident as resolved, saying that traffic was healthy again.
Users also reported a possible Google Gemini outage Thursday morning. However, Google did not confirm a Gemini disruption or list an incident on its service status dashboard. The company did not respond to WIRED’s request for comment before publication.
Multiple outages affecting AI services at the same time can point to a problem with a shared cloud provider, content delivery network, or other third-party infrastructure vendor. In this case, however, OpenAI and Anthropic did not identify a common cause. Major internet infrastructure providers, including Cloudflare, Amazon Web Services, and Microsoft Azure, also reported no outages on Thursday.
Source: www.wired.com