Nobody Is Saying Why OpenAI and Anthropic Had Outages Today

September 3, 2026 | Source: Wired AI

Tags: OpenAI, Anthropic, xAI, ChatGPT, Claude, Grok, SpaceX, infrastructure outage

ChatGPT, Claude, and Grok went offline simultaneously on September 3 — OpenAI blamed a routing error (resolved in 34 min), Anthropic stayed silent on root cause, and xAI pointed to its Memphis compute center, implicating SpaceX's shared compute infrastructure given its partnerships with both Anthropic and xAI.

Details

Frontier AI services from OpenAI, Anthropic, and xAI all experienced outages within a narrow 80-minute window on September 3, fueling speculation about shared infrastructure dependencies. Anthropic's systems went down first at 6:23am PT, with elevated errors on Claude Mythos 5.1, Claude Fable 5.1, and Claude Opus 5; the company resolved the issue by 9:16am PT but declined to explain the cause. xAI's Grok followed at 6:30am PT, explicitly blaming an outage at its Memphis compute center and noting it would 'apologize to our impacted compute partners.' OpenAI's ChatGPT and Codex were hit by a routing error starting at 7:43am PT, with service restored by 8:17am PT. The likeliest common thread is SpaceX: Anthropic and xAI both announced compute partnerships with SpaceX in May 2026, and xAI's reference to 'compute partners' in its apology strongly suggests the Memphis outage may have cascaded. Neither Anthropic nor OpenAI confirmed a shared cause, and AWS, Azure, and Cloudflare all reported no issues of their own. For teams running production workloads on these APIs, the incident reveals that multi-provider redundancy strategies may not fully hedge against correlated failures if the underlying compute infrastructure overlaps — a risk not always disclosed in provider SLAs.