Elon Musk publicly reversed his earlier criticism of Anthropic on Thursday after users on X suggested he might one day disconnect the AI lab from SpaceX servers to hobble a rival. Musk replied with praise, saying such a move would be “not my style.” He also wrote that he had been wrong about Anthropic, referencing a September 2025 X post in which he said: “Winning was never in the set of possible outcomes for Anthropic.”
Even at that time, reports indicated Anthropic already held the largest AI market share among enterprise customers.
The hosting deal and its terms
By July 2026 Anthropic had become one of SpaceX’s largest customers. In May Anthropic signed a deal to buy 300 megawatts of compute capacity — the entire output of xAI’s Colossus 1 data center near Memphis, Tennessee. (xAI merged with SpaceX in February.)
Anthropic agreed to pay $1.25 billion per month through May 2029, a contract that represents roughly $40 billion in revenue for SpaceX’s xAI unit over its term. Separately, Google signed an agreement to rent SpaceX infrastructure as well, paying $920 million per month through June 2029.
Musk’s stated rationale and precedents
Musk insisted the decision was not dangerous for Anthropic and expressed admiration for the rival. He wrote that Anthropic is “obviously currently the leader in AI,” noting no company has released a model as strong as Mythos/Fable and predicting Mythos 2 will arrive soon. He added he would never cut off a competitor in a way that would severely hurt them, calling such action “not my style.”
To illustrate his approach, Musk pointed to Tesla’s 2014 decision — outlined in a now-deleted blog post and preserved under Tesla’s patent pledge — not to initiate patent lawsuits against parties who in good faith want to use its technology. He also noted Tesla opened its Supercharger network and charging port design to competitors. Musk wrote that SpaceX launches competing satellite systems without raising prices or using unfair terms, and that even his worst enemies can still attack him on his platform.
Risks, contractual protections and technical visibility
The article emphasizes Anthropic need not rely solely on Musk’s stated “style.” A sudden shutdown of Anthropic’s infrastructure would carry contractual consequences. There are also strong incentives for SpaceX to keep the deal intact: Anthropic provides substantial revenue, and SpaceX engineers can gain experience building and supporting Anthropic-scale AI systems, similar to how Amazon engineers work with large AI customers.
Hosting Anthropic’s compute could give SpaceX greater operational visibility into how Anthropic builds and runs its models — a potentially valuable vantage point. During Musk’s trial against OpenAI, he acknowledged that AI “distilling” — creating many fake accounts to probe a competitor’s model to learn its behavior — is real. The New York Times reported that, when asked if xAI had ever distilled technology from OpenAI, Musk replied: “Generally AI companies distill other AI companies.”
In February Anthropic accused three Chinese model makers of distilling its Claude model. Presumably Anthropic and Google believe they have safeguards against SpaceX engaging in similar practices while using its infrastructure, but hosting can nonetheless increase SpaceX’s insight into Anthropic’s operations.
Immediate gains, long-term uncertainty
At present the partnership looks overwhelmingly beneficial to SpaceX: high, steady revenue and opportunities to build technical expertise. But the contract runs for several years, and as it ages the balance of incentives and the potential for disputes or strategic tensions could change. Who controls access, what protections remain effective, and how commercial or legal dynamics evolve will matter as the multi‑year relationship unfolds.
