← Latest briefing

Technology

OpenAI, Anthropic, and xAI experience concurrent service disruptions

Major artificial intelligence chatbots suffered simultaneous downtime on Thursday morning, though providers cited distinct internal causes.

The short version

  • Leading AI services including ChatGPT, Claude, and Grok suffered concurrent disruptions on Thursday morning before resolving them.
  • OpenAI pointed to a routing error, while SpaceX reported an infrastructure issue at its Memphis data center affecting Grok.
  • Anthropic addressed elevated errors across multiple Claude models but declined to elaborate publicly on the underlying cause.
  • It remains unconfirmed whether a shared infrastructure dependency played a role in the overlapping timing of the incidents.

Key facts

  • OpenAI reported that a routing issue starting at 7:43 a.m. PT made ChatGPT and Codex unavailable for some users until a fix was applied around 8:17 a.m. PT.[Wired]
  • Anthropic detected elevated error rates impacting Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5 starting at 6:23 a.m. PT, resolving the incident by 9:16 a.m. PT.[Wired]
  • SpaceX stated that Grok's service interruptions, which began around 6:30 a.m. PT and resolved by 10:05 a.m. PT, were caused by an outage at its Memphis compute center.[Wired]
  • Major cloud and network infrastructure providers, including Cloudflare, Amazon Web Services, and Microsoft Azure, did not report outages on Thursday.[Wired]
  • Google did not report any incidents on its status dashboard despite scattered user reports of issues with Gemini.[Wired]

What remains uncertain

  • Whether the concurrent outages stemmed from a single shared upstream vendor dependency or were merely coincidental remains unconfirmed by the companies.[Wired]
  • Whether Google Gemini experienced actual operational downtime has not been verified or acknowledged by Google.[Wired]

Sources