SpaceX has acquired xAI, in a move aimed at integrating AI, launch systems, satellite internet and direct-to-mobile communications under a single corporate structure, the company said.
The acquisition brings xAI’s AI research and products into SpaceX’s broader space and satellite operations, including Starlink and Starship. The transaction values SpaceX at $1 trillion and xAI at $250 billion, Reuters reported.
As part of the deal, xAI investors will receive 0.1433 shares of SpaceX for every xAI share they hold, the report added. Some xAI executives may choose to take cash instead of SpaceX stock, at a price of $75.46 per share, it noted.
In a statement outlining the rationale for the deal, SpaceX said the combined entity would focus on scaling AI compute beyond Earth, arguing that terrestrial data centres face growing constraints on power availability, cooling requirements and environmental impact. “Global electricity demand for AI simply cannot be met with terrestrial solutions,” the statement said.
The company said its long-term solution is to deploy AI compute infrastructure in orbit, using near-constant solar power. “In the long term, space-based AI is obviously the only way to scale,” the statement said, adding that satellites operating as orbital data centres could bypass many of the limits faced on Earth.
According to SpaceX, the plan involves launching a large constellation of satellites that would function as space-based data centres. The company said that with launches every hour carrying about 200 tonnes per flight, its Starship vehicle could eventually deliver millions of tonnes of payload to orbit annually. “Launching a constellation of a million satellites that operate as orbital data centres is a first step,” the statement said.
SpaceX was already the world’s most valuable privately held company, having been valued at about $800 billion. xAI was last valued at $230 billion in November, the Wall Street Journal reported.
The merger comes as SpaceX is planning a public offering later this year that could value the company at more than $1.5 trillion.
SpaceX said that in 2025, despite being the most active year for orbital launches, only about 3,000 tonnes of payload reached orbit, largely through Falcon launches carrying Starlink satellites. The company said Starship would change that equation, starting with the launch of higher-capacity V3 Starlink satellites and next-generation direct-to-mobile satellites providing global cellular coverage.
The company estimated that launching one million tonnes of satellites per year, with each tonne generating 100 kilowatts of compute power, would add about 100 gigawatts of AI compute capacity annually. It said there is “a path to launching 1 terawatt per year from Earth,” and claimed that within two to three years, “the lowest cost way to generate AI compute will be in space.”
SpaceX also said the proposed AI satellite constellation would follow existing space sustainability practices, including end-of-life disposal methods used in its current broadband satellite systems.
Beyond Earth orbit, the company outlined longer-term plans involving lunar and interplanetary operations. It said Starship’s in-space refuelling and heavy-lift capabilities could enable large-scale cargo landings on the Moon, leading to manufacturing facilities that use lunar resources. “Factories on the Moon can take advantage of lunar resources to manufacture satellites and deploy them further into space,” the statement said.
According to SpaceX, lunar manufacturing combined with electromagnetic mass drivers could eventually enable the deployment of AI satellites into deep space. The company said this capability would support scientific research, space-based manufacturing and the establishment of permanent human settlements beyond Earth.
“The capabilities we unlock by making space-based data centres a reality will fund and enable self-growing bases on the Moon, an entire civilisation on Mars, and ultimately expansion to the Universe,” the company said.
ALSO READ: GitHub Introduces Copilot SDK to Embed AI Agents in Applications
