SpaceX Admits Orbital AI Data Centers May Lack Viability
SpaceX disclosed in pre-IPO filings that its orbital AI data centers face technical complexities and may never be commercially viable, contradicting Elon Musk's public optimism.
Space Exploration Technologies Corp. admitted in a confidential S-1 pre-IPO filing that its plans for orbital AI data centers involve significant technical complexity and unproven technologies that may not achieve commercial viability. The disclosure contradicts public statements made by CEO Elon Musk in January 2026, when he described space-based AI as a "no-brainer" and predicted it would become the lowest-cost location for AI within three years. The company warned investors that the unpredictable space environment could lead to system failures and noted that the Starship vehicle has not yet performed as expected to support the project.
Despite these risks, SpaceX has applied to the Federal Communications Commission to operate up to one million satellites for the system. The company is pivoting toward an AI-first business model, using revenue from its Starlink satellite broadband business—which saw operating income double to $4.42 billion last year—to fund massive capital expenditures. In 2025, its AI division, including the acquired xAI, accounted for 61% of total capital spending, totaling $20.74 billion, while reporting an operating loss of $6.4 billion.
SpaceX is targeting a June IPO road show to raise $75 billion at a valuation of $1.75 trillion. The company also reportedly holds an option to acquire the AI startup Cursor for $60 billion. However, some sources dispute these details, claiming there is no verified IPO filing or active program for a million-satellite compute constellation.
