NextFin News - Amazon’s satellite division, recently rebranded as Amazon LEO, formally petitioned the Federal Communications Commission on March 6 to dismiss a radical proposal by SpaceX to launch a constellation of one million satellites dedicated to orbital data processing. The filing, which became public on Tuesday, characterizes Elon Musk’s vision for a space-based AI compute layer as a "speculative placeholder" designed to preemptively seize orbital real estate rather than a viable engineering project. By seeking to occupy nearly every usable altitude in low-Earth orbit (LEO), Amazon argues that SpaceX is attempting to establish a permanent "orbital monopoly" that would effectively lock out all future competitors.
The scale of the SpaceX application, submitted on February 2, is unprecedented in the history of spaceflight. While approximately 15,000 satellites currently orbit the planet, SpaceX envisions a fleet nearly 70 times that size. The company’s thesis rests on the operational maturity of its Starship launch system, claiming it could eventually orbit one million tonnes of hardware annually. According to SpaceX filings, these satellites would generate 100 gigawatts of AI compute capacity, powered by solar energy and freed from the cooling and land-use constraints that plague terrestrial data centers. Musk’s firm asserts that within years, the lowest cost for AI generation will be found in the vacuum of space.
Amazon LEO, which is currently racing to deploy its own Project Kuiper satellites to meet FCC-mandated launch milestones, sees the move as a direct existential threat. In its petition, Amazon noted that SpaceX has failed to provide critical technical details, including specific satellite designs, disposal procedures, or a credible plan to manage "conjunctions"—the industry term for collisions—within a million-unit fleet. The filing suggests that even with SpaceX’s current launch cadence, deploying such a constellation would "take centuries," implying the application is a legal maneuver to "land-grab" orbital shells before other nations or companies can claim them.
The regulatory skirmish has already spilled into the halls of Congress. During a Senate Commerce Committee hearing in February, Senator Maria Cantwell expressed alarm over the "SAT Streamline Act," a bill authored by Senator Ted Cruz that would have granted automatic approval to satellite applications if the FCC failed to act within 18 months. Critics argued the bill would have allowed SpaceX’s million-satellite plan to bypass rigorous safety reviews by default. Following intense debate, the "shot clock" provision was stripped from the legislation, signaling a growing appetite in Washington to scrutinize the rapid privatization of LEO.
The technical skepticism is not limited to corporate rivals. OpenAI CEO Sam Altman recently described the economics of the proposal as "ridiculous," noting that the latency and power-to-weight ratios of current satellite technology do not yet support the massive workloads required for frontier AI models. Furthermore, astrophysicists like Peter Plavchan have warned that a constellation of this magnitude would reach the "carrying capacity" of near-Earth space, creating a physical barrier that could prevent any other actor from launching satellites into those specific orbits. For the FCC, the decision now rests on whether to treat SpaceX’s filing as a visionary leap or a regulatory overreach that threatens the long-term sustainability of the space economy.
Explore more exclusive insights at nextfin.ai.