Anthropic now depends on a competitor's infrastructure for a huge share of its compute. Musk's public reassurance doesn't change the contractual reality: SpaceX has genuine leverage over one of Anthropic's largest cost centers.
- Elon Musk publicly reversed his September 2025 skepticism about Anthropic, calling it “obviously currently the leader in AI” after users suggested he could cut the company off SpaceX’s servers.
- Anthropic pays SpaceX’s xAI unit roughly $1.25 billion per month through May 2029, worth about $40 billion in total revenue, for the entire output of xAI’s Colossus 1 data center.
- Musk cited Tesla’s patent pledge and open Supercharger network as evidence he doesn’t use infrastructure leverage to hurt competitors.
- Musk previously lost a lawsuit against OpenAI, and separately acknowledged under oath that AI companies generally distill each other’s models.
- Google separately pays SpaceX $920 million per month for compute through June 2029.
What Happened
Elon Musk publicly praised Anthropic on Thursday after users on X suggested he could cut the AI lab off from SpaceX’s servers to hurt a rival, TechCrunch reported. Musk called this “not my style” and said he was “clearly wrong” in a September 2025 post questioning whether Anthropic could win.
Anthropic signed a deal in May to buy the entire output of xAI’s Colossus 1 data center near Memphis, Tennessee, paying roughly $1.25 billion per month through May 2029. The deal is worth about $40 billion in total revenue for SpaceX’s xAI unit, which merged with SpaceX in February.
Musk pointed to Tesla’s 2014 patent pledge and its open Supercharger network as evidence he doesn’t weaponize infrastructure against competitors. He also cited SpaceX launching competing satellite systems without raising prices on rivals.
Google separately pays SpaceX $920 million per month for compute through June 2029. Both deals mean two of the largest AI labs now depend partly on a competitor’s infrastructure.
Why It Matters
Musk’s reassurance carries less weight than the underlying contract does. Anthropic and Google have real legal protections and massive financial incentives keeping SpaceX from disrupting service, regardless of Musk’s public statements about his personal style.
Musk’s track record complicates his own framing. He lost a lawsuit against OpenAI after suing a rival directly, and separately acknowledged under oath that AI companies generally distill competitors’ models, the same practice Anthropic accused three Chinese labs of in February and Alibaba of more recently. Hosting Anthropic’s compute gives SpaceX’s engineers a level of operational visibility into Anthropic that most competitors never get.
“That’s not my style.” Elon Musk, CEO, SpaceX and xAI
I was clearly wrong about Anthropic. They are obviously currently the leader in AI. No company has released a model as good as Mythos/Fable and they will undoubtedly have Mythos 2 ready soon.
And I would never cut them off in a way that hurt them badly, even as a competitor.…
— Elon Musk (@elonmusk) July 9, 2026
Bottom Line
Watch how this three-year contract holds up as it matures, particularly if Anthropic and SpaceX’s competitive interests ever directly collide. Public goodwill from an executive is not a substitute for contractual protection.
For SaaS founders relying on any AI lab as a vendor, this is a reminder to look past public statements and understand your vendor’s own infrastructure dependencies. If your AI provider depends on a competitor’s compute, that relationship’s health is now part of your own vendor risk.
