The decade Europe bet its sovereignty on a single technology

For a decade, the question that haunted Europe's quietest policy rooms was never whether the continent could build the future, but whether it would be allowed to. From Brussels to Berlin, from a wind-scoured stretch of the North Sea to a cleanroom outside Grenoble, a single conviction has hardened into doctrine: that a bloc which cannot make its own chips, run its own clouds, store its own energy and feed its own factories is not, in any meaningful sense, sovereign at all. Over six months we followed the money, the machines and the politics behind that conviction — and found a gamble far larger, and far more fragile, than its architects admit in public.

The numbers alone are vertiginous. Counting national subsidies, EU instruments and the private capital they are meant to unlock, more than €312 billion has been pledged across Europe's industrial strategy through 2030. Roughly a third is aimed at a single sector — advanced semiconductors — on the theory that whoever controls the chip controls the car, the data centre, the missile and the heat pump alike. "We spent thirty years believing the market would sort it out," a senior official in the Commission's industry directorate told us, asking not to be named because the framing remains politically delicate. "Then we watched a single Taiwanese company become more strategically important to our economy than most of our member states."

That company, of course, is the one nobody in Europe could replicate, and the realisation arrived as a series of shocks rather than a single epiphany. A pandemic that idled assembly lines in Wolfsburg and Turin for want of components the size of a fingernail. An energy crisis that exposed how much of the continent's prosperity had been quietly underwritten by cheap imported gas. A scramble for the batteries, magnets and refined metals that the green transition would consume by the megatonne — almost all of them processed, today, in a country that is also Europe's chief geopolitical rival. Each crisis pointed at the same uncomfortable conclusion. Europe had optimised brilliantly for a world that no longer existed.

The cathedral in the cornfield

Nowhere is the bet more literal than on the outskirts of Dresden, where a consortium is raising what its engineers call, only half-joking, a cathedral. The fabrication plant — a "fab," in the industry's clipped argot — will eventually cost more than ten billion euros, draw as much power as a mid-sized city and employ several thousand people to make chips at process nodes that did not exist when its foundations were poured. Walk its half-finished corridors, as we did in March, and the scale of the wager becomes physical. The cleanroom alone is the length of three football pitches; the machines that will fill it are so precise that a stray speck of dust is a catastrophe and so expensive that a single one costs more than a commercial aircraft.

What makes the project a gamble rather than a guarantee is everything the steel and silicon cannot buy. A fab is only as good as the engineers who can coax yield out of it, and Europe spent two decades letting that expertise drift to Asia. It depends on a supply chain of ultra-pure chemicals, specialty gases and lithography optics so concentrated that a flood, a fire or an export ban anywhere along it can stall the whole enterprise. And it must compete, year after year, against rivals whose states have been subsidising the same race for far longer and with far deeper pockets. "Building the fab is the easy part," one of its directors admitted. "The hard part is still being competitive in 2032, when the subsidy runs out and the politics have moved on."

That distinction — between dominance and insurance — runs through every conversation we had. Almost no one we interviewed believes Europe will out-manufacture Asia or out-spend the United States. The honest case for the strategy is narrower and, its defenders argue, more durable: that a continent with even a partial domestic capability cannot be cleanly coerced. A bloc that makes ten per cent of its own leading-edge chips, refines its own critical minerals and can island its own grid is harder to squeeze than one that makes none. Sovereignty, in this reading, is not autarky. It is the price of a seat at the table when the table turns hostile.

Energy is the hidden chapter

The semiconductor story dominates the headlines, but spend time with the people executing the strategy and a quieter truth surfaces: none of it works without electricity, and lots of it. A modern fab is among the most power-hungry buildings ever constructed. The AI data centres that the chips are meant to feed are hungrier still — a single large training run can consume as much electricity in a month as a small town does in a year. The same forecast that has Europe building fabs has it racing to build the generation and the grid to run them, and it is here, far from the cleanrooms, that the sovereignty bet is most exposed.

For all the political capital spent on where chips are made, far less has been spent on where their power will come from. Offshore wind, the supposed backbone of the plan, has stumbled on inflation, permitting and a transmission grid that was designed for a different century. Nuclear, rehabilitated after years in the political wilderness, takes a decade and a fortune to build. Batteries and grid-scale storage — the technologies that would let intermittent renewables actually keep a fab running through a windless January night — remain heavily dependent on supply chains the continent does not control. "You can have the most advanced chip in the world," one grid operator told us, dryly, "and it is worth nothing if you cannot keep it powered."

The minerals question is the one that keeps strategists awake. A fab needs silicon, but a battery needs lithium, nickel and graphite; a wind turbine needs rare-earth magnets; a grid needs copper by the millions of tonnes. Europe sits on more of these resources than its imports suggest, but mining them means reopening a politics the continent had thought it left behind — permits that take a decade, communities that do not want a quarry in the valley, and an environmental ledger that the green transition was supposed to clean up, not complicate. You can read our companion data investigation into the regions betting their future on a lithium boom, and the residents pushing back.

A wager on patience

What unites the chip fab, the wind farm and the half-permitted mine is time. The European bet is, at its core, a wager that democratic states can hold a strategy steady across electoral cycles, budget crises and the next inevitable downturn — that the political class will keep writing cheques for a decade before the first dividend arrives. It is a bet against the grain of how European politics has worked for a generation, and its defenders know it. "Markets reward quarters; sovereignty rewards decades," the Commission official said. "We are asking democracies to think like engineers. That is the real experiment here, not the silicon."

Whether it pays off will not be clear for years, and possibly not until the crisis it is meant to insure against actually arrives. For now, the cranes keep turning over the cornfield outside Dresden, the turbines keep rising in the grey North Sea, and a continent that spent thirty years trusting the market to provide is learning, expensively and in public, what it costs to provide for itself. The architects of the strategy speak of it in the language of insurance. Their critics call it a bet. They are both right — and a decade from now, Europe will know which word was the more honest. Read the full series, The Sovereignty Files, as it unfolds.