Anthropic Is Renting AI Compute from SpaceX's Data Center — Elon Musk Now Powers Claude

Anthropic, widely marketed as the 'ethical' AI company, just struck a deal to rent compute from a massive SpaceX-run data center — meaning Elon Musk's infrastructure now keeps Claude running. The irony is hard to miss: users who chose Claude partly for its principled stance on AI development are now, in effect, thanking Musk every time the chatbot responds quickly.

The move is a direct reversal of Anthropic CEO Dario Amodei's earlier posture. He had subtly criticized OpenAI's Sam Altman for aggressively signing compute deals — and is now doing exactly the same thing. The reason is simple: AI demand is absolutely crushing supply. Anthropic has suffered outages and imposed usage limits in recent months, frustrating the developers it most needs to keep happy. On stage at the company's developer conference, Amodei owned up to it: "We've had difficulties with compute. We're sorry if sometimes it takes some time, but we're gonna keep going to acquire as much as we can."

The stakes are real. Some developers had already started migrating to OpenAI's Codex — specifically because OpenAI's earlier compute deals meant fewer restrictions on usage. Shortly after the SpaceX deal was announced, Anthropic relaxed several rate limits, particularly around Claude Code, its fast-growing AI coding product. That's a direct competitive response.

The backdrop to all this is staggering growth. Anthropic is reportedly expanding at roughly 80x annual revenue pace. Business Insider reporter Stephen Council, covering his first Anthropic developer conference, captured what that looks like from the inside: features are flipping from research preview to public beta almost overnight, new hires are giving demos after just weeks on the job, and the company is shipping so fast that even its own workers say they can barely keep up. At that growth rate, sustained for even another year, Anthropic would become one of the highest-revenue companies on the planet.