Elon Musk's SpaceX—merged with xAI earlier this year—is team up with Anthropic to power Claude, establishing an unusual partnership between two companies that previously opposed the administration of President Donald Trump.
On Wednesday, announced that Anthropic will gain access to Colossus 1, a massive AI supercomputer cluster built by xAI for training and running advanced AI models. Anthropic stated that this agreement will provide it with over 300 megawatts of additional computing power and help enhance service for Claude Pro and Claude Max users.
Over the next few days, we will accelerate the operation of Claude’s reasoning system on the Colossus detector. Thank you for partnering with SpaceX,” said Tom Brown, co-founder of Anthropic. Wrote X: “To meet the demands of artificial intelligence, we need to move vast amounts of atoms, and no one is better at moving atoms quickly—whether on Earth or beyond.”
In subsequent posts about X, Musk said that conversations with senior staff at Anthropologie helped convince him that the company is developing artificial intelligence responsibly.
“Everyone I’ve met has been extremely competent and deeply committed to doing the right thing,” Musk wrote. “No one triggered my evil detector. As long as they engage in critical self-reflection, Claude is probably a good person.”
Musk added that "SpaceXAI" has already shifted training to Colossus 2, a larger next-generation AI supercomputer cluster designed to surpass existing systems. The Colossus 1 supports approximately 220,000 NVIDIA GPUs and enables future Grok models.
This partnership comes at a time when Musk and Anthropic’s CEO Dario Amodei are increasingly representing different camps within the AI industry. Musk has aligned himself closely with Trump and the U.S. Department of Justice, defending while Anthropic positions itself as one of the industry’s strongest advocates for AI safety standards. positioned
This stance recently brought the company into conflict with the Trump administration and the Pentagon, after Anthropic refused requests to allow unrestricted military use of the Klaud drone. The controversy centers on how to implement measures to prevent this drone model from being used for large-scale domestic surveillance or the development of fully autonomous weapons. Trump later ordered federal agencies to phase out Anthropic’s technology, while Pentagon officials argued that military contractors must allow the government “any lawful use” of artificial intelligence systems.
The protocol highlights how the competition for computing power is reshaping Silicon Valley's alliance landscape. In February, Elon Musk folded xAI to question SpaceX, arguing that power demands and cooling requirements limit the growth of AI data centers on Earth. Clarified X announced on Wednesday that the xAI entity will be dissolved, and this portion of the business will henceforth be called “SpaceXAI”.
Also in March, Musk was accused of being "misogynistic" after Anthropic raised $30 billion in funding, reaching a post-money valuation of $380 billion.
Despite Antropic's previous strong statements, he said that this collaboration with SpaceX further strengthens the company’s growing infrastructure partnership, which now includes Amazon, Google, Broadcom, Microsoft, and NVIDIA.
As part of this agreement, we also expressed interest in collaborating with SpaceX to develop gigawatts of orbital AI computing capacity,” Anthropic stated.
Although the orbital data center is still in the experimental stage and faces significant engineering challenges, SpaceX has stated that ground-based power, cooling infrastructure, and available land may not scale quickly enough to meet the demands of future AI systems, prompting interest in space-based computing as a potential source of larger, more sustainable energy supplies.
The company stated: "If engineering challenges can be overcome, space-based computing could provide nearly limitless sustainable energy with a smaller impact on Earth."