Europa, from client to competitor of big tech?
First warnings in the Old Continent to chart an autonomous path in artificial intelligence. The knots to be untied are still many, perhaps too many. And the pushes from various countries do not go in a single direction, with France trying to sprint ahead.
| 21 June 2026
Artificial intelligence is racing. And it speaks with a Californian accent. The dominant models come from the United States, cloud infrastructures see a growing Asian presence, and European companies find themselves building their technological future on often external foundations. In this scenario, the question is no longer theoretical: can Europe still play an autonomous game in AI, or is it destined to become an advanced user of other people's technologies? According to various analyses by the EU Commission, Europe's delay in artificial intelligence is not just a question of "how much is invested," but of "how" it is invested and, above all, how quickly research can be transformed into business. Added to this is another critical element: the difficulty in building a complete supply chain, from semiconductors to computing power to models and industrial applications, still largely dependent on non-European actors.
For this reason, the European Commission has just put two new strategies on the table with a declared objective: to bring AI into industry and turn it into the engine of European scientific research. The line is drawn by Ursula von der Leyen: Europe wants a "made in Europe" artificial intelligence, capable of combining innovation, security, and technological autonomy. Not just principles, but an operational plan that aims to rapidly spread AI in key sectors – from healthcare to energy, automotive to manufacturing – and to strengthen the European position in algorithm-driven science. On the industrial front, Brussels pushes for concrete adoption: one billion euros to support businesses and public administration, with particular attention to small and medium-sized enterprises, often excluded from the most advanced technologies.
The leap in scale, however, requires much more significant figures. With the InvestAI initiative, the European Union aims to mobilize up to 200 billion euros between public and private funds, including about 20 billion for the construction of a limited number of artificial intelligence "gigafactories," high-capacity computing infrastructures for the development of the most advanced models. Among the measures, AI-based health centers, development of frontier models and "vertical" applications for industry, environment, and pharmaceuticals. In support, an alliance for applied AI will be created, bringing together businesses, universities, and institutions, while an observatory will monitor the impact of technology in different sectors. In parallel, Brussels also tries to simplify the regulatory front with a dedicated desk for the implementation of the AI Act, whose entry into force will take place gradually until 2027.
European Commission President Ursula von der Leyen wants a "made in EU" artificial intelligence (© Getty Images)
The second leg of the strategy looks at science. Here the ambition is even higher: to make Europe a global hub for research based on artificial intelligence. The core is Raise, a virtual institute that will coordinate resources, data, and expertise at the European level. Significant resources are coming: 600 million to enhance computing capacity, dedicated access to future AI gigafactories, and a doubling of research investments to over 3 billion annually. Also on the table are funds to attract talent and strengthen doctorates, with the aim of retaining in Europe skills increasingly contested globally. In parallel, initiatives like InvestAI aim to mobilize public and private capital on a large scale, also for the construction of large computing infrastructures and data centers.
The structural knot, however, remains: data. For this reason, Brussels is preparing a new strategy for the Data Union, expected shortly, which will have to guarantee access to quality information, today fragmented and often unused. It is on this ground that a decisive part of the global competition is played, together with the availability of energy and infrastructure necessary to support the large-scale development of AI.
The United States accuses Europe of "regulating too much and innovating too little." A criticism that Anne Le Hénanff, French delegate minister for Artificial Intelligence and Digital, does not completely reject, but reinterprets: "The challenge is not to choose between innovation and regulation, but to find the right balance." The reference is to the AI Act, the first attempt in the world to organically regulate artificial intelligence. "We must progress towards a union of capital markets," Le Hénanff emphasizes. Only then will it be possible to finance the growth of start-ups and tech scale-ups.
© Getty Images
Another critical point is infrastructure. Without European cloud and computing capacity, talking about digital sovereignty remains difficult. Projects like Gaia-X have tried to respond, but with still limited results compared to the speed of global competitors. In the meantime, however, signs of reaction are beginning to emerge: some countries, starting with France, are accelerating with national plans that provide for total investments exceeding 100 billion euros between AI and data centers, while startups like Mistral AI have raised over 1.7 billion euros, becoming one of the few European candidates to compete in the race for frontier models.
Meanwhile, European companies continue to use services offered by large non-EU providers. A choice often inevitable in the short term, but which raises strategic questions in the long term. In Montreal, the climate between Europe and the United States was "good," but marked by latent tensions. Even episodes like Elon Musk's reaction to European sanctions show how the regulatory ground has become a field of geopolitical confrontation. Europe claims the right to set its own rules – from the Digital Services Act to the AI Act – and to enforce them. But the decisive confrontation remains with the two technological superpowers. The question, in the end, is as simple as it is crucial: does Europe want to be a protagonist or a spectator?
One Continent, three approaches
In Europe, the adoption of artificial intelligence does not follow a single model, but a plurality of trajectories that reflect different economic priorities, political visions, industrial capabilities, and investment capacities. Three main approaches can be identified. The first is the "industrial" one, led by France and Germany, where AI is seen as a lever of competitiveness in key sectors – manufacturing, automotive, energy – with significant public investments and support for the emergence of technological champions. The second is the "infrastructural" one, adopted by Spain and the Nordic countries. In this case, the priority is to build the foundations: supercomputing, data management, cloud, and digitization of public administration, with AI as a system enabler. The third is the "regulatory" one, more evident in European institutions and partly in Italy, where attention focuses on governance, security, and protection of rights, to develop reliable artificial intelligence in line with European values. These models are not alternative but complementary. The knot remains the ability to integrate them: without industrial scale and adequate capital, the risk is that the EU remains an excellent regulatory and scientific laboratory, but dependent on technologies developed beyond its borders.
More than technological, it is a financial lag
In the United States, venture capital flows with a depth and speed that is difficult to replicate: a few large, highly capitalized funds can sustain long investment cycles, burn resources for years, and accompany companies to global scale. It is this mechanism that has allowed the birth and expansion of the big AI players. In Europe, on the contrary, venture capital remains more fragmented, less risk-prone, and often limited by not fully integrated financial markets. The result is that many promising startups fail to make the leap in scale or end up being acquired by foreign groups precisely in the crucial growth phase. The second lag is that of industrial scalability.
Europe excels in basic research – universities, centers of excellence, public and private laboratories produce very high-level innovation – but struggles to build industrial platforms capable of competing on a global scale. In the United States, the ecosystem combines universities, private capital, and big tech in a virtuous circle; in China, the state plays a direct role, mobilizing massive resources on infrastructure, data, and applications. Europe, instead, moves more slowly: still segmented national markets, regulations not always harmonized, and a lower availability of critical infrastructure – starting with computing capacity and large data centers – make it more difficult to go from experimentation to large-scale deployment.
This double fragility is reflected in the daily choices of companies. Even when innovation is born in Europe, it is often developed or commercialized elsewhere. Thus, many European companies – from large corporates to small businesses – end up integrating into their processes models developed by American big tech, exploiting platforms that offer computing power, development ecosystems, and access to data that are difficult to replicate at home. On the infrastructural front, the use of non-European cloud services is often a forced choice rather than a strategic one: competitive costs, reliability, and immediate scalability push towards global providers. In the short term, this dependence allows European companies to remain competitive and to quickly adopt advanced solutions. But in the medium-long term the risk is deeper: shifting value abroad, loss of control over data and critical technology chains, less ability to direct innovation according to European industrial priorities. In other words, it is not just a technological issue, but one of economic sovereignty. If the fundamental tools of the digital economy are designed, updated, and controlled elsewhere, even strategic choices – from industry to security – inevitably end up being influenced by those who develop those systems.
This article is part of the third edition of the AI Special, published in the June 2026 issue of Business People. Download the issue or subscribe here
Stay always updated with the new Business People Whatsapp channel
© Reproduction reserved
