Once upon a time, a company would build large, costly, and unwieldy satellites that would then be sent into orbit to sit, often for decades. Those satellites did their job, supporting navigation, tracking the weather, and enabling disaster response, as they circled the planet.
These days, they do much more. They are crucial to modern armies, for example, since they underpin command and control, precision targeting, secure links, early warning, logistics, and intelligence. And their environment, orbit, has also changed. It is contested. The head of the UK’s Space Command says British satellites are targeted by Russia weekly.
Why satellites must change faster than humans can manage
It is unsurprising, then, that satellites are changing rapidly. They used to be the size of buses; now, they are often the size of basketballs. The twin pressures of the market and technological change have forced the hand of satellite operators, who must refresh their hardware more often.
That hardware must serve multiple missions, steer beams at breakneck speed, and respond to shifting demand. What holds them back now is not the technology, but the human beings who increasingly struggle to allocate satellite capacity at the speed required. Agile constellations, multi-mission systems, fluctuating demand: the current approach will not do.
The global zeal for artificial intelligence has been so wildly out of proportion with the reality of what it can actually do that a theorised AI bubble has spawned its own Wikipedia page. But AI does, of course, have its uses, and it has already made a difference in a large number of sectors. Space is one of them.
Satellites need to detect patterns, update plans in close to real time, and adjust behaviour within an environment that is partly structured. This has all the hallmarks of a classic AI use case. Applied to capacity management, AI could decide who gets bandwidth, when, and for how long. In contested settings, this matters a great deal. It can also improve resilience, since automated systems can keep networks running even when links are degraded or operators are overloaded.
Moving intelligence from the ground to orbit
AI will also transform how satellites handle data. Today, the bulk of the filtering and interpretation of data collected by satellites takes place on the ground. Earth observation firms, for instance, which might track the weather, crop health, disaster response, land use, or emissions, must merge and clean vast streams of imagery to derive the insights their users need.
If satellites can process the bulk of that data in orbit and downlink only what can be used, the cost savings will be huge. In fields such as defence, where decision-making is a key differentiator, individuals will be able to take action far more quickly than they otherwise could. And although satellites, as already noted, are not immune to harm, whether accidental or deliberate, they are less vulnerable than their counterparts on the ground.
Downstream of this are further gains in innovation. Whenever machines take over from humans, space is created for those humans to do something else. Indeed, there is sometimes a transition period, during which operators need to develop new skills or companies need to reorganise themselves, but technology-driven efficiency ultimately increases innovation.
Machines can simulate creativity, but they cannot, at least not yet, be creative. Only human beings can. With AI handling orchestration and on-board data processing, engineers and operators can focus elsewhere, and innovation becomes more likely.
Europe’s structural challenge
What this means is that AI startups could play an outsized role in deciding who wins the space race. That should be of particular interest to policymakers in Europe, which, despite leadership in the development of so-called human-centric AI, lags behind the US and China in artificial intelligence.
Adoption has been slower, markets are fragmented, and the focus has been heavily weighted towards regulation, as demonstrated by the passing of the AI Act. Europe’s hand may be forced if the US and China, already pulling ahead in the space race, make full use of AI in their space operations. Given the absolute centrality of space to so many areas of human life, from finance to logistics to defence, Europe may be compelled to change tack.
That will probably require a cultural shift. As is often noted, the way Europe operates as a bloc has historically prioritised fairness over brute efficiency. Funding is fragmented, procurement cycles are slow, early-stage ventures struggle to scale, and effective public-private collaboration in space and defence remains in its infancy.
A more pragmatic approach, one that focuses on the challenges Europe faces today and responds accordingly, may be needed. There are signs that this is already changing, but in truth, the pace is not fast enough.
For now, the reality is that space is changing, and AI startups are likely to accelerate that change. The governments that back those startups, and their populations, will benefit.
Read the orginal article: https://www.eu-startups.com/2026/02/the-space-race-is-being-rewritten-by-ai-and-europe-risks-falling-behind/
