The public assurance follows a massive infrastructure agreement finalized in May, in which Anthropic secured the entire 300-megawatt output of the Colossus 1 data center near Memphis. The deal, valued at $1.25 billion monthly through May 2029, positions Anthropic as a primary customer for SpaceX’s xAI unit. Google has also entered the fold, renting infrastructure for $920 million per month through June 2029. Musk, who once claimed Anthropic had no path to victory, now calls the company the current leader in artificial intelligence, citing their Mythos and Fable models as industry benchmarks.
In section Startups & Technology
Elon Musk pledges stability for Anthropic’s massive compute deal
“I was clearly wrong about Anthropic,” Elon Musk admitted Thursday, publicly walking back his previous skepticism toward the AI lab. Despite lingering industry anxiety over whether Musk might leverage his control of SpaceX infrastructure to kneecap a competitor, the billionaire insists he remains committed to his high-stakes partnership with the firm.
To deflect concerns about potential sabotage, Musk pointed to his historical track record of opening Tesla’s charging network and patent portfolio to rivals. However, the arrangement carries inherent risks. While Anthropic is protected by strict contractual obligations, the physical proximity of their operations to SpaceX’s engineering teams creates a unique environment. Musk has previously acknowledged that AI "distilling"—the practice of probing a competitor's model to reverse-engineer its capabilities—is a standard industry tactic. With SpaceX engineers gaining direct insight into Anthropic's scaling needs, the partnership offers Musk’s company significant, if unconventional, advantages beyond the $40 billion revenue stream.
Comments (0)
No comments yet. Be the first!