Elon Musk’s SpaceX has announced that it has signed a deal with Anthropic to facilitate access to XAI’s massive hardware to increase the computing capability for its Claude AI models. The announcement was made when Anthropic held a developer day in San Francisco, where it unveiled its new feature “dreaming”, which is meant to help AI systems work between sessions, spot patterns, update files that store user preferences and other context.
As part of the deal, Anthropic gains exclusive access to the full capacity of the Colossus 1 supercomputer in Memphis, Tennessee. It is home to more than 220,000 Nvidia processors and will give the Claude chatbot maker 300 megawatts of new capacity within a month. The Colossus edge is currently the largest supercomputer in the world and will help Anthropic run Claude at a higher “compute density” than they could on standard cloud providers like AWS or Google Cloud. It also “expressed interest” in working with SpaceX to develop multiple gigawatts of compute capacity in space.
Anthorpic will essentially pay $3–$4 bn annually. At the same time, this puts XAI at par with other computing giants like Google, Microsoft or AWS, and having a revenue stream from its computing service, it makes XAI more valuable to Wall Street investors. Musk further gets “the kill switch” as a political and ethical leverage over one of its biggest competitors, which can monitor, gatekeep, the operational scale, load, and efficiency of an AI giant.
Musk has always criticised Anthropic; he said that it was “doomed to become the opposite of its name,” which would be misanthropic. He called the company “hypocritical”, and now, to some extent proving his point. In February 2026, Musk said that Anthropic “hates Western Civilisation”. But on Wednesday, he shifted his rhetoric; he said that he was “impressed” after meeting the team.
In a social media post, Musk wrote, “Everyone I met was highly competent and cared a great deal about doing the right thing. No one set off my evil detector.” He added, “So long as they engage in critical self-examination, Claude will probably be good.”
The two AI model positioned themself as polar opposite to each other. Anthropic’s Claude AI thumps its chest over Constitutional AI (a written “bill of rights” for the AI), which Musk dismissed as “enforced diversity”. The Anthropic has hypocritically put itself into the same orbit as Elon Musk’s “anti-woke” and “maximalist” philosophy for xAI. Anthropic claims that it’s just renting the engine while hands steering the wheel remain with itself. Whereas Musk have appointed himself as the arbitrator and reserves the right to reclaim the compute if the AI engages in actions he deems “harmful to humanity.” While analysts suggest both of them are in bed together and continue their performative rivalry and consolidate power by controlling the air that AI models need to breathe.