SpaceXAI apologizes for outage that affected Grok and other ‘compute partners’
SpaceXAI has apologised after an outage at its Memphis data centre knocked its Grok chatbot offline for more than three hours on Thursday, also affecting unspecified "compute partners". The disruption struck Grok across X and its Android and iOS apps, and comes as the firm increasingly supplies computing capacity to other artificial intelligence companies, making the incident notable beyond its own userbase.
Grok's outage began at 6:30am PT and was logged on the company's status page as a "models outage" before systems were restored. SpaceXAI did not explain the cause, though Elon Musk said corrective action was being taken to prevent a repeat. The timing coincided with separate problems at Anthropic, whose Claude models suffered elevated errors from around 6:30am PT until resolution at 9:16am PT, and at OpenAI, where ChatGPT and Codex saw elevated errors from roughly 7:30am PT until 9:55am PT. Anthropic leases compute from SpaceXAI under a deal signed earlier this year, but neither company has confirmed whether the incidents were linked, and both declined to comment further.
- SpaceXAI's Memphis data centre outage took Grok offline for over three hours
- Anthropic's Claude and OpenAI's ChatGPT also hit errors around the same time
- Link between the incidents remains unconfirmed; all services since restored