Why Is Google Taking AI Data Centres Into Orbit, and What Should UK Businesses Do in 2026?
Google is exploring AI data centres in orbit, and the reason will be familiar to anyone watching the UK grid: power is the limiting factor. For British businesses, the interesting question is not rockets but what energy-bound AI does to costs.
What is the Concept
Google calls the research Project Suncatcher: solar-powered satellites carrying its AI chips, connected by lasers, running machine learning workloads in space. A prototype launch with a partner is planned for around 2027, so it is an experiment rather than a product.
For a UK SME the point is straightforward. The biggest providers are searching for energy, and that affects capacity and pricing in the regions you rely on, including London.
Why It Matters Now (2025–2026 Context)
The UK government has designated AI Growth Zones and named data centres as critical national infrastructure, yet grid connection delays remain a well-known obstacle, particularly around London and the South East.
The contrarian view: orbital compute will not shrink your sterling cloud bill any time soon. The common mistake is treating AI spend as a rounding error until the invoice arrives.
How AI Is Changing This
AI agents multiply model calls per task. A Manchester accountancy practice that automates document review across hundreds of clients can see usage climb quickly once staff trust the tool.
The non-obvious idea is to budget AI like energy: set allowances per team and review them monthly.
Real-World Examples
Large projects such as the planned Stargate UK and expansions around Slough and Newport show demand for domestic compute is growing, and so is scrutiny of its power use.
A realistic scenario: a Leeds e-commerce firm uses AI for product descriptions. Routing routine items to a smaller model and using a larger one for premium lines keeps quality up and monthly GBP spend down.
Practical Insights / Actions
Use the Compute Resilience Ladder: measure, route, cap, diversify. Know your cost per outcome, choose the cheapest capable model, cap agent loops, and keep an alternative provider ready.
Check compliance as well. Confirm where prompts are processed, document your lawful basis under UK GDPR, and ask vendors for data processing terms before scaling.
Future Outlook
Orbital data centres face launch, radiation, heat and servicing challenges. Expect most UK capacity to remain on the ground, supported by renewables, grid upgrades and more efficient chips.
The hidden opportunity is efficiency. British firms with lean AI workflows will cope better if energy costs keep pushing compute prices up.
Conclusion
Google's orbital plan signals that AI is energy-limited. Respond by measuring and trimming your own consumption. RP SoftTech can help UK organisations audit AI workloads and put practical cost controls in place.
Frequently Asked Questions
Is Google really building data centers in space?
Google has described Project Suncatcher, a research effort to run AI chips on solar-powered satellites. Prototype launches are planned around 2027, so it is still experimental.
Will space-based AI data centers lower my cloud costs?
Not in the near term. Launch, cooling and maintenance are hard problems, so most AI capacity will stay on Earth this decade. Optimising current usage saves more now.
What is the fastest way to cut AI compute costs in the UK?
Measure cost per task, route simple work to smaller models, cap agent loops and add caching. These steps often reduce spend without hurting quality, and need no new vendor.
How should SMEs plan for AI infrastructure constraints?
Track spend by workflow, avoid long contracts before measuring usage, keep a second provider tested, and ask vendors about capacity and pricing if power supply tightens.