On September 3, 2026, several leading AI chatbot services suffered widespread outages affecting users globally. OpenAI's ChatGPT and Codex, Anthropic's Claude models, xAI's Grok, and Google's Gemini all faced disruptions during the day, causing major interruptions across platforms [1, 2, 3, 4, 5, 6, 7].

OpenAI traced its ChatGPT and Codex outage to a routing error that began around 7:43 am PT (14:43 UTC) and resolved the issue by 8:17 am PT (15:17 UTC). An OpenAI spokesperson, Kathleen Chaykowski, said, "A routing error starting around 7:43 am PT on Thursday, September 3, made ChatGPT and Codex unavailable for some users across platforms. As of about 8:17 am PT on Thursday, a solution was successfully implemented and is continuing to be monitored" [4, 7].

Anthropic reported elevated error rates and partial outages for multiple Claude models including Mythos 5.1, Fable 5.1, and Opus 5. The root cause was infrastructure-related, and fixes were deployed by early afternoon UTC. According to Anthropic status updates, "We have identified the cause of elevated errors affecting Claude models and have deployed a fix. The incident was resolved by 9:16 am PT and we continue to monitor" [2, 4, 7].

xAI's Grok experienced outages starting early morning PT caused by issues at their Memphis compute center. The company said, "Grok is experiencing issues. We are working on restoring service as quickly as possible. We have resolved the situation, and traffic is healthy again." Service was restored by 10:05 am PT [4].

Google's Gemini chatbot faced user-reported issues on September 3, but Google did not officially confirm any outage or update its status dashboard. While some users and media noted fluctuations in Gemini's service, the company remained silent on the matter [4, 5, 8, 6, 7].

Downdetector recorded over 12,000 user problem reports for ChatGPT, between 1,200 and 6,000 reports for Claude, and approximately 1,000 for Grok on September 3 [2, 3, 9]. Some reports linked the outages to shared cloud infrastructure dependencies, particularly Microsoft Azure, which reportedly faced concurrent service issues [8, 6, 7].

The simultaneous outages across several major AI platforms on one day mark a rare event, causing widespread service disruptions worldwide [2, 8, 6, 7].

Timeline of key events on September 3:

  • 6:23 am PT: Anthropic alerts users to partial outages with elevated errors on several Claude models [4].
  • 7:43 am PT: OpenAI encounters a routing error disabling ChatGPT and Codex for some users [4].
  • 8:17 am PT: OpenAI fixes ChatGPT and Codex issues [4].
  • 10:05 am PT: xAI reports Grok outage resolved [4].
  • 12:16 pm ET (9:16 am PT): Anthropic announces Claude services repaired [7].
  • 12:55 pm ET: OpenAI reports ChatGPT and Codex fully restored [7].

Companies continue to monitor their AI services after the outages on September 3.