Elon Musk has publicly softened his stance toward Anthropic, saying he would not deliberately disrupt the AI company's access to infrastructure. His comments followed online speculation about whether he could restrict the lab's use of SpaceX resources.
Musk said he had been "clearly wrong" about Anthropic and praised its model performance, calling it the current leader in AI. He also suggested that a next-generation version of its flagship model could arrive soon, underscoring the pace of progress in the sector.
The shift comes as Anthropic remains one of SpaceX's major customers. The company reportedly secured a large compute agreement in May for access to 300 megawatts of capacity from xAI's Colossus 1 data center near Memphis, Tennessee. The arrangement is set to run through May 2029 and represents a major long-term infrastructure commitment.
Musk pointed to Tesla's patent pledge and the opening of its Supercharger network as examples of how his companies have historically approached competition. He also said SpaceX has launched competing satellite systems without imposing unfair terms.
Beyond the business relationship, the partnership highlights how closely AI development and compute infrastructure are now linked. For SpaceX, hosting major AI workloads can bring revenue, technical insight, and deeper involvement in the next phase of intelligent systems. The broader trend suggests that future AI growth may increasingly depend on strategic infrastructure alliances.