
If approved, SpaceX’s orbital compute network could dramatically lower AI infrastructure costs and reshape the data‑center industry, while raising regulatory and spectrum management challenges.
The FCC’s invitation for public comment marks a rare regulatory window into what could become the most ambitious data‑center project ever imagined. Historically, the commission has required phased deployments to ensure spectrum efficiency and avoid interference. By asking for a waiver of the six‑year half‑deployment and nine‑year full‑deployment milestones, SpaceX is signaling confidence in Starship’s rapid launch cadence and a desire to accelerate the timeline far beyond conventional satellite constellations. Stakeholders—from telecom operators to AI startups—will be watching the comment period closely, as the outcome could set precedent for future mega‑constellations.
SpaceX’s vision hinges on leveraging reusable launch vehicles to loft a million tonnes of satellite hardware, each delivering roughly 100 kW of compute power. At scale, this translates to 100 gigawatts of AI processing capacity, potentially undercutting terrestrial data‑center costs driven by land, power, and cooling expenses. The orbital environment also offers near‑continuous line‑of‑sight to global users, reducing latency for certain AI workloads. However, the technical challenges are non‑trivial: radiation hardening, thermal management, and on‑orbit servicing will dictate actual performance and lifespan. If SpaceX can solve these engineering hurdles, the economics could shift AI compute from a land‑bound commodity to a space‑based utility.
Regulators and industry analysts are weighing the broader implications. Spectrum allocation for a million satellites raises concerns about interference with existing services and the need for robust coordination mechanisms. Environmental groups are likely to scrutinize the orbital debris risk, especially given the sheer volume of hardware. Meanwhile, competitors such as Amazon’s Project Kuiper and traditional hyperscale providers may need to reassess their own infrastructure strategies. The FCC’s final decision, expected after the March comment cycle, will not only determine SpaceX’s roadmap but also signal how quickly the market can transition to space‑centric AI compute.
You have a month to comment · February 10, 2026 · Sebastian Moss
The Federal Communications Commission said that it is seeking public comment on SpaceX's orbital data‑center plan.
FCC Chairman Brendan Carr said the agency “welcomes” the proposal to launch a million data‑center satellites.
Image: SpaceX satellite – Getty Images
SpaceX claims that, once Starship is operational and reusable, “launching one million tonnes per year of satellites generating 100 kW of computer power per tonne would add 100 gigawatts of AI compute capacity annually, with minimal ongoing operational or maintenance needs.
“Freed from the constraints of terrestrial deployment, within a few years, the lowest cost to generate AI compute will be space.”
No specific timelines were shared. SpaceX requested a waiver of FCC milestone requirements that usually require half of a constellation to be deployed within six years of authorization and the full system within nine years.
First, the plan requires FCC approval, with the agency calling for public comments until 6 March. It plans to respond to those by 16 March, with the public again able to respond to those comments by 23 March.
Comments
Want to join the conversation?
Loading comments...