United Nations has repeatedly highlighted that space applications underpin progress across all 17 Sustainable Development Goals.What’s missing is not ambition or innovation, but a coordinated international ecosystem that enables scale while ensuring accountability. Against this backdrop, space‑based data centres (SBDCs) are emerging as both an opportunity and a test of how ready the global system really is.
Access to space has long been tied to national economic and security priorities. Telecommunications, Earth observation, research, and downstream data services already play a direct role in economic growth and technological development. But the infrastructure that supports these capabilities has become increasingly constrained on Earth, this is most noticeable with data centers.
Global data centres consumed an estimated 415 terawatt-hours of energy in 2024, more than the total electricity usage of some countries. Additionally, a single data centre can require hundreds of thousands of gallons of water per day for cooling. These pressures are forcing operators to look beyond terrestrial solutions. SBDCs seem to promise a compelling alternative. By operating in orbit, they could sidestep many of the bottlenecks faced on the ground such as lengthy utility approvals, right‑of‑way negotiations, environmental impact assessments, and power availability constraints. A single launch could replace years of permitting and construction, opening the door to faster deployment and rapid scaling. SBDCs could potentially generate electricity at $0.002/kWh, which is about 22x cheaper than current terrestrial costs, and it does so by using the 95% solar availability in a sun synchronous orbit.
However, there both regulatory and technical challenges that need to be addressed. Much of the current debate around SBDCs centres on cooling. Cooling in space is unintuitively complex, while space is freezing cold, the reality of cooling objects is more nuanced. Without an atmosphere, heat cannot be transferred through convection and requires radiation.That means heat management depends on carefully designed radiators, often positioned away from direct solar exposure.
Additionally, a growing body of evidence shows that frontier‑grade AI chips don’t last nearly as long as traditional hardware. The performance demands are so intense and the pace of innovation so fast that these accelerators face heavy thermal and electrical stress. According to the Princeton Center for Information Technology Policy, that combination of wear‑and‑tear and rapid obsolescence can shrink their effective lifespan to just one to three years.
If the technical challenges are addressed, governance has not kept pace. Large constellations of space objects raise familiar but unresolved concerns including space debris, orbital congestion, emissions from launch activity, and long‑term sustainability. On the geopolitical side, questions of sovereignty, strategic dominance, and monopolisation remain highly sensitive.
There is also commercial reality. Capital‑intensive infrastructure favours large incumbents. If only a handful of actors can deploy and operate SBDCs at scale it could serve only to further entrench divides among developed and developing nations.
From a legal perspective, existing treaties were never designed with orbital cloud infrastructure in mind. Issues such as data sovereignty, cross‑jurisdictional operations, liability, enforcement, and the growing privatisation of outer space remain unresolved in current NGSO activities already, let alone further developments of SBDC. These gaps do not block innovation today, but they will shape who benefits from it tomorrow as countries and regions decide how to regulate this cross-jurisdictional technology.
Despite these challenges, SBDCs represent a genuine paradigm shift. They offer the possibility of solar‑powered, in‑orbit AI processing, extending cloud infrastructure beyond Earth and relieving pressure on terrestrial grids. This matters because demand is accelerating fast. By 2030, AI training clusters alone could consume close to 9% of US electricity, levels that existing grids are unlikely to sustain.
It’s no coincidence that many have publicly flagged energy as the limiting factor for next‑generation AI. Space‑based compute reframes the problem entirely moving power generation and processing closer together, and outside the constraints of national grids.
This vision is no longer speculative. In January 2026, SpaceX filed an application to launch and operate an orbital data centre system, proposing a constellation of up to a million satellites.[1] The filing highlights near‑constant solar power, low operating costs, and a reduced environmental footprint as key advantages, while noting that global data‑centre electricity demand is expected to more than double by 2035.
Space‑based data centres sit at the intersection of sustainability, geopolitics, and next‑generation computing. However, it remains to be seen if they become a catalyst for global progress or if they amplify existing inequalities and systemic risks.
Which path prevails will depend less on technology and more on anticipatory governance. Environmental safeguards, political cooperation, and legal clarity must evolve alongside commercial ambition. Space has long been treated as a shared domain, and preserving its accessibility while enabling innovation is a challenge the international community can no longer postpone.