SpaceX has acquired Elon Musk’s artificial‑intelligence start‑up xAI and announced an audacious plan to build “space‑based data centres” powered by solar energy captured in orbit. In a memo posted on SpaceX’s website, Musk framed the move as a response to the growing energy and cooling demands of terrestrial AI infrastructure and presented a long‑term strategy that ties AI, Starlink broadband and Starship launches into a single vertically integrated engine of innovation.
Bloomberg reported that the combined company would be valued at about $1.25 trillion; Reuters has previously noted that roughly 80% of SpaceX revenues come from Starlink, and xAI is said to be burning roughly $1 billion a month. Musk did not address an imminent IPO in his public statement, though SpaceX has reportedly been preparing for a flotation as early as June. The acquisition folds together two of Musk’s high‑profile but cash‑hungry ventures and raises immediate questions about financing, governance and cross‑subsidy between commercial businesses with distinct missions.
Musk’s technical pitch is unapologetically grand: he argues that ground‑based data centres consume untenable amounts of electricity and that shifting compute to orbit would let satellites harvest near‑constant solar power at low operating cost. He envisages very large constellations — eventually measured in the hundreds of thousands or millions of satellites — launched by increasingly capable Starship vehicles, and lays out back‑of‑the‑envelope arithmetic that links per‑ton compute density, annual launch mass and gigawatts of added AI capacity.
The proposal is as bold as it is speculative. Moving high‑density computing into orbit confronts hard engineering problems: collecting and converting solar energy at scale, managing waste heat in vacuum, hardening hardware against radiation, and building high‑throughput inter‑satellite and ground links with low latency. There are also industrial constraints: manufacturing, testing and launching millions of satellites would require dramatic ramp‑ups in supply chains and a sustained Starship cadence far beyond what the sector has achieved to date.
Regulatory, environmental and reputational issues compound the technical hurdles. The U.S. Federal Communications Commission requires end‑of‑life deorbit plans; orbital sustainability and debris mitigation will be central to any regulatory approvals. xAI’s recent controversies — including criticism over a Tennessee data centre and reports that the Grok chatbot was used to produce abusive imagery — add reputational risk to the deal. National security actors will scrutinise any large‑scale, dual‑use satellite compute capability, and spectrum allocation, export controls and allied concerns could shape how quickly such a project can proceed.
From a business viewpoint, the acquisition creates both synergies and liabilities. If SpaceX can monetise an orbital compute layer it would secure a durable, high‑margin revenue stream that complements Starlink connectivity and justifies continued Starship investment. But the short‑term arithmetic is difficult: xAI’s cash burn and the capital intensity of large constellations mean that investors and regulators will demand clarity on financing and on how responsibilities and risks are shared across Musk’s companies.
This is ultimately a long‑term vision rather than an immediate roadmap. Delivering the promised economics will take years of technological maturation, regulatory sign‑offs and capital deployment. Even so, the move underscores a growing strategic convergence: whoever controls low‑cost compute and ubiquitous connectivity will have an oversized influence on the next phase of AI development, with consequences for commercial competition, national security and space governance.