SpaceX to Provide Anthropic Access to AI Supercomputer

SpaceX is set to give Anthropic access to its massive artificial intelligence supercomputer, according to Reuters, in a compute infrastructure deal between two prominent U.S. technology companies.

The arrangement would expand Anthropic’s computational resources beyond its existing cloud computing arrangements, according to Reuters. The deal pairs Elon Musk’s aerospace and technology company with the AI safety-focused laboratory founded by former OpenAI executives, including CEO Dario Amodei.

The arrangement comes amid intense competition for AI compute capacity among leading laboratories. Anthropic, the maker of the Claude family of AI models, has relied primarily on cloud partnerships — most notably with Amazon Web Services, which has committed billions of dollars in investment in the company — to power its model training and inference workloads.

SpaceX’s entry into the AI compute market represents an expansion of the company’s technology footprint beyond its core launch and satellite businesses. The move aligns with Musk’s broader AI ambitions, which also include his separate AI venture xAI, maker of the Grok chatbot.

The deal underscores the growing strategic importance of compute infrastructure in the AI industry, where access to large-scale GPU clusters has become a critical bottleneck for model development. Major AI labs have pursued diversified compute strategies, seeking capacity from cloud providers, custom data centers and now aerospace firms with significant computing assets.

For Anthropic, the arrangement could reduce its dependence on any single infrastructure provider while bolstering the computational resources needed to train increasingly large foundation models. The company has been competing with OpenAI, Google DeepMind and other labs in a race to develop more capable AI systems.

Financial terms of the arrangement were not immediately disclosed in the Reuters report.