SpaceX Acquires xAI in Move to Integrate AI and Starlink

Elon Musk announced that SpaceX has acquired his artificial intelligence startup xAI, combining the two companies into what he described as “the most ambitious, vertically integrated innovation engine on (and off) Earth.” The merger integrates SpaceX’s launch capabilities, its Starlink satellite infrastructure, and plans for lunar operations with xAI’s artificial intelligence technologies, with the goal of scaling compute capacity through space-based data centers.

“This marks not just the next chapter, but the next book in SpaceX and xAI’s mission,” Musk said in a statement, outlining a long-term vision that spans artificial intelligence, rockets, space-based internet, direct-to-mobile communications, and a real-time information platform.

The merger comes amid surging global demand for AI compute resources, which Musk cited as a key driver for relocating compute-intensive operations to space. “Current advances in AI are dependent on large terrestrial data centers, which require immense amounts of power and cooling. The only logical solution, therefore, is to move these resource-intensive efforts to a location with vast power and space,” Musk wrote.

Musk and SpaceX envision a constellation of up to 1 million satellites functioning as orbital data centers, harnessing near-constant solar energy to meet the energy demands of AI workloads. “By directly leveraging near-continuous solar power with minimal operating and maintenance costs, these satellites will transform our ability to scale compute,” Musk said.

According to filings with the Federal Communications Commission, SpaceX has requested approval to deploy such a constellation, aimed at accommodating the rapid growth in data demand driven by AI. If SpaceX achieves its planned launch cadence, Musk said, each satellite would contribute to a cumulative AI compute capacity of up to 100 gigawatts annually.

“The lowest cost way to generate AI compute will be in space within two to three years,” Musk added, predicting that orbital infrastructure will allow companies to train models and process data at unprecedented scale.

SpaceX’s Starship  launch vehicle is central to this strategy. The company expects Starship to eventually reach launch rates of up to one per hour, with each flight capable of carrying roughly 200 tons of payload. Such capacity would enable the deployment of millions of tons of satellites annually, laying the foundation for space-based compute infrastructure. Musk has also outlined longer-term applications for Starship, including cargo delivery to the Moon and the establishment of permanent scientific and manufacturing bases.

While orbital AI data centers represent the near-term objective, Musk emphasized the Moon as a critical site for future manufacturing. “Once there, it will be possible to establish a permanent presence for scientific and manufacturing pursuits,” he said. Lunar facilities could leverage local resources to manufacture satellites, which could then be launched deeper into space using electromagnetic mass drivers. Musk suggested such an approach could ultimately generate between 500 and 1,000 terawatts per year of AI compute capacity from orbit.

The proposed constellation would build on SpaceX’s existing satellite sustainability practices, including end-of-life disposal protocols developed through the Starlink program. As of 2025, Starlink operates more than 9,000 satellites, with continued expansion planned through its next-generation V3 satellites. Each V3 launch is expected to deliver more than 20 times the capacity of current V2 satellites, further strengthening SpaceX’s global broadband and direct-to-device connectivity offerings.

The SpaceX-xAI merger also reflects a broader trend among technology companies seeking new sources of compute to support AI development. Analysts note that AI workloads are driving sharp increases in data center electricity demand, which Goldman Sachs projects could rise by 165% by 2030. Jensen Huang, CEO, Nvidia, has similarly warned that next-generation AI models may require up to 100 times more compute power than earlier systems. Musk’s space-based approach aims to meet these needs while reducing the environmental and community impacts associated with terrestrial data centers.

Financially, the merger positions SpaceX and xAI for an initial public offering. Bloomberg has reported that the combined entity could pursue an IPO at a valuation of roughly US$1.25 trillion. SpaceX was valued at approximately US$800 billion following a secondary share sale in December 2025, while xAI’s was valued at US$230 billion after a funding round in January 2025. 

Despite the strategic alignment, cultural integration may present challenges. xAI has historically operated with a fast-moving, flat organizational structure, while SpaceX is known for its rigorous engineering discipline and operational hierarchy. “I have a hunch many xAI people will hit culture shock with SpaceX,” said Benjamin De Kraker, a former xAI staffer. Some personnel already hold dual roles across Musk’s companies, including Christopher Stanley, who serves as a principal security engineer at SpaceX and senior director of security engineering at X, Musk’s social media platform integrated with xAI.

The combined SpaceX–xAI platform further extends Musk’s broader corporate ecosystem, which also includes Tesla and X, formerly known as Twitter. Musk has repeatedly emphasized the strategic value of cross-company synergies in addressing large-scale challenges. “This new constellation of satellites will enable innovative companies to forge ahead in training AI models and processing data at unprecedented speeds and scales,” he said, “accelerating breakthroughs in our understanding of physics and the invention of technologies that benefit humanity.”

Industry observers note that SpaceX’s move into AI compute infrastructure could set a precedent for space-based commercial applications. Deploying a constellation of one million satellites would represent an unprecedented logistical and technological undertaking, potentially dwarfing historical launch records. In 2025, the most active year on record, roughly 3,000 tons of payload were delivered to orbit. Musk envisions Starship eventually enabling deliveries of millions of tons per year, supporting continuous satellite deployment and longer-term ambitions for lunar and Martian infrastructure.

Musk has framed the initiative within his broader vision of a multi-planetary future. The combination of orbital AI data centers, lunar manufacturing, and eventual expansion to Mars is intended to enable what he describes as Kardashev Type II–level capabilities—harnessing solar energy at a scale far beyond current Earth-based systems. “Thanks to advancements like in-space propellant transfer, Starship will be capable of landing massive amounts of cargo on the Moon,” Musk said. “The capabilities unlocked by space-based data centers will fund and enable self-growing bases on the Moon, an entire civilization on Mars, and ultimately expansion into the universe.”