In a bold move that would have set the stage for a new space race, OpenAI's Chief Executive Sam Altman explored assembling funds to either acquire or partner with a rocket company, directly positioning himself against Elon Musk's SpaceX. This strategic exploration, which has since stalled, underscores the growing intersection of artificial intelligence ambitions and space infrastructure.
The Stoke Space Proposal and Its Stalling
According to sources familiar with the discussions, Altman reached out to rocket manufacturer Stoke Space in the summer of 2025, with talks intensifying during the fall. A key proposal involved OpenAI making a series of equity investments in the company, potentially totalling billions of dollars over time, to eventually secure a controlling stake. However, people close to OpenAI have confirmed that these talks are no longer active.
Stoke Space, founded by former employees of Jeff Bezos' Blue Origin, is developing a fully reusable rocket named Nova. Acquiring a stake would have given Altman direct exposure to this technology, entering a market where SpaceX holds a dominant position in launch services. This would have further intensified the existing rivalry between Altman and Musk, who also competes through his AI startup, xAI.
Driving Vision: AI Data Centers in Space
The potential rocket investment was not merely about launch capabilities; it was tied to a grander vision for AI's future. Altman has publicly expressed interest in building data centers in space to power next-generation artificial intelligence systems. The rationale is that the insatiable computing demands of AI could eventually have significant environmental consequences on Earth, making orbital data centers powered by solar energy a viable alternative.
"I do guess that a lot of the world gets covered in data centers over time," Altman said on a podcast with Theo Von. "Maybe we build a big Dyson sphere around the solar system and say, 'Hey, it actually makes no sense to put these on Earth.'" This concept, while unproven, has also been discussed by other tech leaders like Google's Sundar Pichai.
Market Headwinds and OpenAI's Current Challenges
These exploratory talks took shape during a peak in market enthusiasm for AI. In September and October 2025, Altman announced major chip and data center deals with companies including Oracle, Nvidia, and Advanced Micro Devices, promising a vast build-out of computing infrastructure. However, the market sentiment has since cooled. Oracle shares have fallen about 19% in the last month, while Nvidia declined some 13%.
OpenAI itself is navigating significant pressures. The company has committed to nearly $600 billion in new computing deals in recent months, raising questions about its financing strategy, especially as it projects $13 billion in revenue for this year. Internally, the company declared a "code red" to improve ChatGPT after losing market share to Google's Gemini, delaying other product rollouts and reassigning staff to focus on the core chatbot.
Furthermore, Nvidia's CFO revealed this week that its $100 billion deal with OpenAI has not yet been finalized. Altman, a veteran venture capitalist with a portfolio of over 400 companies, has not shied away from using OpenAI's balance sheet for ambitious projects, such as a recent $18 billion commitment to a data-center venture called Stargate with SoftBank.
The proposed foray into rocketry, had it proceeded, would have marked another frontier in Altman's expanding empire, which already includes brain-computer interface startup Merge Labs (a Neuralink competitor) and a social network project. For now, the dream of an OpenAI-backed rocket to challenge SpaceX and enable orbital AI clusters remains grounded, a testament to the immense technical, financial, and regulatory challenges of the final frontier.
