Where do they grow?

Europe produces researchers, publications, patents, universities and startups. Then, too often, it looks the other way when those startups have to become large.

In artificial intelligence the distance is obvious. In 2024 US institutions produced forty AI models regarded as internationally significant, against fifteen from China and just three from Europe as a whole. In the same year US private investment in artificial intelligence reached 109.1 billion dollars, a figure vastly higher than that of the other major economic areas.

Europe's problem is not the absence of intelligence. It is the absence of scale.

Building an advanced model requires capital, energy, computing power, data, researchers, engineers and a commercial structure able to turn research into a product. It requires accepting years of investment and losses before reaching a market. It requires investors capable of understanding deep technologies and of bearing risks that an ordinary financier would hardly accept.

Europe is often good at the early stage: public research, incubation, grants, prototypes. It becomes weaker when the company has to hire hundreds of people, buy infrastructure, enter several markets quickly and raise large rounds of capital.

According to the European Investment Bank, European scaleups raise on average roughly half the capital obtained by their Silicon Valley equivalents. The fragmentation of financial markets and the scarcity of large specialised investors push many companies to look for capital, listings or buyers outside the Union.

The point is not that no rules exist in the United States. There are rules, litigation, taxation and bureaucracy. But the system has deeper capital markets, a large and relatively unified economic space, and a greater cultural willingness to accept aggressive growth, failure and the temporary concentration of resources on a few promising firms.

Europe, instead, often asks a startup to behave like a mature company before it has even proved it can survive.

The result is a disproportion. A young firm with ten employees may face compliance duties, consultancy, differing national regulations, labour rules, tax systems and capital-access procedures that absorb an enormous share of its organisational capacity. For a multinational those costs are a line item; for a startup they can be the difference between growing and closing.

A study published in January 2026 by the European Commission and the European Investment Bank identified, among the main reasons for the relocation of European startups, easier access to venture capital, proximity to large and unified markets, more favourable regulatory environments, and the availability of commercial talent with international experience. Often it is not the whole company that moves: the registered office and management shift abroad, while research and engineering stay in Europe. In some cases it is the US investor itself that requires the move.

It is a peculiar form of industrial loss: we keep the laboratory, but we export ownership, control, taxation and decision-making power.

For Italy the difficulty is even greater. On top of European fragmentation come administrative slowness, interpretive uncertainty, scarce risk capital and an ecosystem in which many firms are born thinking first of obtaining a subsidy rather than of winning a market.

Public support can be useful, but it must not turn the entrepreneur into a professional filler-in of grant applications. A startup should not measure its effectiveness by the number of funded projects, but by the number of problems solved for customers willing to pay.

Real innovation policy does not consist in protecting young firms indefinitely. It consists in letting them grow quickly, fail quickly when necessary, and try again without being branded forever.

The European Union has recognised part of the problem. The European strategy for startups and scaleups aims to reduce fragmentation, broaden financing, attract talent and introduce a more uniform European company regime. In March 2026 the "EU Inc." path was launched, conceived as a common set of company rules for innovative firms.

But no reform will work if Europe continues to regard growth in size with suspicion.

A technology company can be born small, but it cannot aspire to stay small. In artificial intelligence, biotechnology, space, energy and semiconductors, scale is not a financial vanity: it is a component of the technology itself.

Digital sovereignty is not achieved merely by writing the rules for products built by others. It is achieved by building products, infrastructure and companies able to honour our values without depending entirely on foreign capital and platforms.

Europe must decide what it wants to be: the place that defines the limits of the future, or the place that, while respecting those limits, also manages to build it.

Why do European ideas set off?

Perhaps because here they can be born.

But they cannot always grow.