In the race between artificial intelligence developers, there is an unusual cooperation between two rivals.
May 7, 2026, 7:08 amMay 7, 2026, 7:08 am
The AI company Anthropic is getting much-needed computing power from a huge data center owned by Elon Musk.
Anthropic buys from Elon Musk.Image: keystone
What makes the partnership surprising: Anthropic is a direct competitor of Musk’s AI developer xAI and the tech billionaire criticized the rival as “anti-human” just a few months ago.
Anthropic announced that it would use the entire capacity of the Colossus 1 data center of Musk’s space company SpaceX, into which xAI has now been incorporated. This means access to more than 220,000 AI chips from Nvidia, the leading provider of semiconductors for artificial intelligence. Anthropic boss Dario Amodei said the company is trying to offer users more computing power as quickly as possible.
Musk’s AI company is merged into SpaceX
Musk also announced that xAI would be dissolved as an independent company and that in future AI products would be available directly from SpaceX under the name SpaceXAI.
xAI developed the chatbot Grok, which competes with Anthropic’s AI software Claude. Both companies, in turn, are rivals of the ChatGPT developer OpenAI, with whom Musk has had an escalating feud for years. A court case is currently underway in which Musk wants to ensure that OpenAI boss Sam Altman loses his job and the competitor’s structure is changed.
Musk now “impressed”
Anthropic was also recently attacked by Musk. In February, he claimed on his online platform X that Anthropic’s AI hated white people, heterosexuals and men. That is “misanthropic and evil”.
Now Musk took a completely different tone. He said he spent a lot of time last week with top Anthropic staff to find out how they were making sure Claude was good for humanity. He was “impressed”. That’s why he agreed to rent Colossus 1 to Anthropic – while his own AI development has now moved to the next data center, Colossus 2.
Data centers in space
SpaceX also said that Anthropic was interested in collaborating on the project to build data centers for artificial intelligence in space. Musk argues that due to energy shortages, among other things, enough computing power for future AI applications can only be placed in orbit. However, there are doubts as to whether such an idea can be implemented profitably.
Anthropic between Pentagon and myth
Anthropic has been in the headlines again and again in recent months. The company risked conflict with the US government by insisting, despite massive pressure, that its AI models could not be used in autonomous weapons systems and for mass surveillance in the US. The Pentagon then declared Anthropic a supply chain risk, which could severely impact the use of the company’s software in government agencies. The company is suing.
Recently, Anthropic demonstrated how important the company’s technology is for the US government. A new Anthropic model called Claude Mythos Preview finds vulnerabilities in various software that have remained undetected for decades. In the wrong hands, it would be a dangerous cyber weapon – and at the same time could give the US government an advantage in hardening its own programs and exploiting potential vulnerabilities for espionage abroad. Anthropic has no plans to make Mythos Preview public. (sda/dpa)