Google and SpaceX are in talks to launch data centers into orbit, according to a Wall Street Journal report dated May 12, 2026. The discussions arrive as SpaceX prepares a $1.75 trillion IPO later this year, with orbital AI compute pitched as a central part of the investor narrative. Google is reportedly speaking with other launch providers as well, hedging its bets on access to space.
The Google side of the conversation is anchored by Project Suncatcher, an initiative the company announced late last year to test compute satellites in low Earth orbit. Google plans to fly prototype satellites by 2027 under that program, which would mark the first concrete step from concept to hardware. The SpaceX talks would give Suncatcher dedicated launch capacity at scale.
SpaceX has its own AI compute strategy already in motion. The company acquired xAI in February 2026 and last week struck a deal with Anthropic to use computing resources from xAI's Memphis, Tennessee data center, with a stated option to collaborate on orbital infrastructure in the future. The Anthropic agreement gave SpaceX a marquee AI customer for its terrestrial capacity and a logical bridge to whatever it builds above the atmosphere.
Key facts
- 01Google and SpaceX are in talks to launch orbital data centers, per a Wall Street Journal report dated May 12, 2026.
- 02SpaceX is preparing a $1.75 trillion IPO later this year, pitching orbital compute as a core thesis.
- 03Google plans to launch prototype satellites by 2027 under Project Suncatcher, announced late last year.
- 04Google invested $900 million in SpaceX in 2015, per regulatory filings.
- 05SpaceX acquired xAI in February 2026 and signed a compute deal with Anthropic last week tied to xAI's Memphis facility.
Google and SpaceX are not strangers commercially. Google invested $900 million in SpaceX in 2015, per regulatory filings, an early bet on Starlink that has since appreciated as SpaceX's valuation climbed toward the trillion-dollar mark. A data center partnership would deepen a relationship that has so far been mostly financial.
“SpaceX is selling investors on orbital data centers as the cheapest place to host AI compute within a few years, with a $1.75 trillion IPO planned for later this year.”— Jaeden Schafer
The economic case for orbital compute rests on three claims: continuous solar power, passive cooling in vacuum, and freedom from the local permitting fights that have slowed terrestrial buildouts in the United States. Elon Musk has been the loudest advocate, claiming orbital data centers will be cheaper to operate than ground-based ones. SpaceX is using a version of that argument to support its IPO valuation.
The counter-argument is straightforward. Once satellite construction and launch costs are factored in, today's terrestrial data centers remain significantly cheaper per unit of compute than anything in orbit. Bandwidth back to Earth, radiation hardening, and servicing constraints add further costs that do not show up in glossy renderings. The economics shift only if launch costs continue to fall sharply, which is itself a SpaceX-dependent assumption.
There is also a political dimension to the timing. Ground-based AI data centers have drawn opposition in several US states over power draw, water use, and tax abatements. Maine recently considered a statewide moratorium before its governor vetoed the bill, and similar fights are underway in Virginia, Georgia, and Texas. Orbital sites sidestep that politics entirely, though they introduce new questions around spectrum, orbital debris, and export controls.
Google's interest also has to be read against its terrestrial commitments. The company recently expanded its TPU partnership with Anthropic and Broadcom into the multi-gigawatt range, and Alphabet has signed power deals stretching years out. Orbital compute is not a replacement for that buildout in any near-term scenario; it is a hedge and a research bet, sized to matter only if launch economics keep improving.
Neither Google nor SpaceX has confirmed the talks publicly, and the Wall Street Journal report cites unnamed sources. Project Suncatcher remains at the prototype stage, and the 2027 launch target is for test hardware rather than production capacity. A serviceable orbital data center business, if it materializes, is a late-decade story at the earliest.
For SpaceX, the strategic value is immediate regardless of physics. Pitching a $1.75 trillion IPO requires a credible story about where the next leg of growth comes from after Starlink saturates, and AI compute is the most expensive line item in the technology economy right now. Tying SpaceX's launch cadence to Google's compute roadmap, and to Anthropic's demand for capacity, gives public-market investors a concrete narrative even if the orbital servers themselves remain years away. The talks are less about 2027 hardware than about anchoring the AI infrastructure trade to launch capacity that only one company currently controls.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
