The CEO of OpenAI has quietly explored acquiring a rocket company to build space-based data centers, potentially creating the first direct competition to Elon Musk’s SpaceX.
Story Snapshot
- Sam Altman held talks to take controlling stake in Stoke Space Technologies for orbital data centers
- Strategic move targets space-based AI infrastructure to challenge SpaceX’s growing dominance
- Negotiations ran through summer 2025 but no deal has been publicly announced
- Represents potential new front in AI infrastructure race extending into space
The Quiet Space Gambit
Sam Altman’s exploration of acquiring Stoke Space Technologies reveals a calculated strategic shift that extends far beyond typical venture investments. The OpenAI CEO reportedly pursued a controlling stake in the Kent, Washington-based rocket startup during mid-2025, with negotiations stretching through the summer and fall. Unlike conventional launch service contracts, Altman’s interest centers on building dedicated orbital infrastructure specifically designed to support massive AI computational demands.
The timing proves particularly revealing. As OpenAI’s models grow exponentially more complex and data-hungry, traditional terrestrial data centers face mounting limitations in power consumption, cooling costs, and physical space constraints. Space-based data centers offer theoretical advantages including abundant solar energy, natural cooling from the vacuum of space, and potential regulatory arbitrage opportunities that could prove crucial for AI development at scale.
Watch:
Stoke Space’s Strategic Value
Stoke Space Technologies represents more than just another rocket startup in Altman’s calculus. Founded in 2019 by former Blue Origin engineers, the company focuses on fully reusable launch systems with rapid turnaround capabilities that could enable the high-cadence missions necessary for orbital data center deployment and maintenance. Their emphasis on reusable second stages and cost reduction aligns perfectly with the economics required for space-based infrastructure projects.
The company’s location in the Seattle aerospace corridor provides additional strategic benefits, offering access to established aerospace talent pools and proximity to major cloud computing operations from Microsoft and Amazon. This geographic positioning could facilitate the complex integration of space-based and terrestrial computing infrastructure that any orbital data center project would require. Stoke’s technical approach to reusability could potentially offer cost structures competitive with or superior to SpaceX’s current capabilities.
Sam Altman eyes rocket company to take on Elon Musk in space race https://t.co/swXFPQxmcG
— FOX Business (@FoxBusiness) December 5, 2025
Challenging the SpaceX Monopoly
Elon Musk’s SpaceX currently enjoys near-monopolistic control over multiple critical space infrastructure layers. Starlink provides global satellite internet coverage, Falcon 9 dominates commercial launch services, and the developing Starship platform promises unprecedented payload capacity to orbit. Any company seeking to deploy significant space-based infrastructure must essentially negotiate through SpaceX’s controlled bottlenecks, creating both cost pressures and strategic dependencies.
Altman’s potential Stoke acquisition represents the first serious attempt by a major AI company to bypass this dependency through vertical integration. Rather than remaining beholden to SpaceX’s pricing and availability, an Altman-controlled rocket company could prioritize AI-specific missions and develop specialized capabilities for orbital data center deployment. This approach mirrors successful vertical integration strategies in terrestrial cloud computing, where companies like Google and Amazon built custom chips and data centers to support their specific computational requirements.
The Broader Infrastructure Chess Game
The Stoke discussions fit within Altman’s broader pattern of pursuing control over fundamental AI infrastructure layers. His previous initiatives have included securing chip manufacturing capacity, exploring nuclear energy partnerships, and building strategic relationships with cloud providers. Space-based infrastructure represents the logical next frontier in this systematic approach to ensuring OpenAI’s long-term competitive position through infrastructure independence.
However, the orbital data center concept remains largely theoretical despite significant technical interest. Current space-based computing consists primarily of small satellite payloads rather than true data centers. The engineering challenges of deploying, maintaining, and networking substantial computational infrastructure in space present formidable obstacles that even well-funded projects have yet to overcome at commercial scale. Success would require breakthrough advances in space-rated computing hardware, automated maintenance systems, and high-bandwidth space-to-ground communications.
Sources:
Seattle Business Journal – OpenAI Stoke Space SpaceX Sam Altman ChatGPT
GeekWire – OpenAI CEO Stoke Space Data Center
IDN Financials – Sam Altman Explores Buying a Rocket Company Competing with Elon Musk
