SpaceX has asked the Federal Communications Commission for permission to launch up to 1 million solar-powered satellites, describing them as orbiting data centers built to meet the surging demand for artificial intelligence computing.
In its filing, SpaceX claims space-based computing could be the “most efficient” way to scale AI while reducing Earth-bound energy constraints.
The proposal goes far beyond near-term infrastructure. SpaceX frames the plan as an early step toward a Kardashev Type II civilization, capable of harnessing a significant share of the Sun’s energy, and as part of its long-term goal of securing humanity’s multi-planetary future.
However, The Verge notes the headline number is unlikely to be approved as-is and may serve as a negotiating position. Recently, the FCC approved 7,500 additional Starlink satellites but deferred authorization for nearly 15,000 more, signaling caution as constellations rapidly expand.
Orbital crowding is already a concern. The European Space Agency estimates around 15,000 human-made satellites are currently in orbit, raising alarms over space debris and light pollution. At the same time, Amazon is seeking an FCC deadline extension for its own satellite network due to rocket shortages, while SpaceX is reportedly exploring a merger involving Tesla and xAI (which has already merged with X) ahead of a potential public listing.
Why it matters: If approved—even at a fraction of the request—SpaceX’s plan could reshape how AI infrastructure is built, shifting massive compute workloads off Earth and into orbit. But regulators now face a balancing act between innovation, safety, and the growing risks of an increasingly crowded sky.





