Europe has made technological regulation one of its defining features. However, increasingly, it’s becoming harder to argue that this alone is enough to compete with the United States and China in the AI race. This fundamental debate is reopening after a new wave of criticism towards the European model: lots of rules, extensive oversight, and still limited independent strength in chips, cloud computing, foundational models, and computing infrastructure.
The issue is far from minor. The EU has indeed gained regulatory influence with the AI Act, which went into effect on August 1, 2024 and will be fully enforceable, with exceptions, starting from August 2, 2026. Brussels has established a framework aiming to regulate AI deployment based on risk levels, with specific rules for prohibited practices, general-purpose models, and high-risk systems. The problem is that regulation does not automatically translate into leadership, especially when much of the critical infrastructure is still imported from outside.
In the tech sector, this tension is increasingly clear: Europe has learned to arbitrate but has yet to demonstrate that it can build at the same pace as its rivals. In AI, where scale matters as much as regulation, this gap weighs more than Brussels would like to admit.
The AI Act grants regulatory power but doesn’t resolve dependency
The AI Act is arguably the EU’s most significant demonstration of regulatory power in this field. It sets transparency obligations, rules for general-purpose models, and restrictions on unacceptable uses. Legally and politically, it’s an important move. Yet, from an industrial perspective, it doesn’t automatically address Europe’s reliance on foreign providers in critical parts of the value chain.
This is the most uncomfortable aspect of the debate. Major cloud platforms, leading models, much of the foundational software, and a significant portion of the hardware that supports advanced AI remain predominantly American-owned or, in various industrial segments, heavily exposed to Asia. Europe can impose access conditions, enforce compliance, and set standards, but that does not mean it controls the strategic assets necessary for the next decade of technological development.
Mario Draghi summarized this more formally in his report on European competitiveness. The report states that Europe’s share of global technological revenues declined from 22% to 18% between 2013 and 2023, while the U.S. share increased from 30% to 38%. Additionally, only four of the 50 largest tech companies worldwide are European. These figures do not indicate a complete defeat but do show a clear relative decline in sectors that will shape the next phase of growth.
Brussels is trying to realign through investment and infrastructure
It would be unfair, however, to portray the EU solely as a regulatory body. Over the past two years, the Commission has actively worked to bolster its industrial capacity. In February 2025, Ursula von der Leyen launched InvestAI, an initiative to mobilize €200 billion in AI investment, including €20 billion allocated to AI gigafactories. The clear message: Europe needs computing power, capital, and scale to transition from a regulated market to a relevant producer.
This effort is complemented by the AI Factories initiative within EuroHPC. The Commission reports that Europe already has 19 AI Factories deployed using supercomputers, along with 13 AI Factory Antennas to expand regional access to this capacity. Furthermore, the community plan envisions new gigafactories focused on training and deploying advanced models. On paper, these are significant steps toward closing infrastructure gaps with the US and China.
There is also a focus on semiconductors. With the European Chips Act, the EU aims to reindustrialize part of the chip ecosystem and reduce vulnerabilities. While not an immediate solution for the AI race, it’s a crucial piece if Brussels wants its technological autonomy to move beyond mere rhetoric toward a coherent industrial policy.
The core issue is not just regulation but strategic independence
The most interesting aspect of the European moment is that criticism has shifted; it’s no longer just about overregulation. The core reproach is that Europe has long confused regulation with strategy. Setting standards can give influence, but it doesn’t make a technological power if you don’t control infrastructure, capital, talent, and execution capacity. That’s where the US and China are playing a different game: one dominates the cloud, software, and much of global venture capital; the other pushes industrial policy, manufacturing, and internal scale. Meanwhile, Europe is now trying to accelerate a transition that ideally should have started earlier.
This doesn’t mean Europe is out of the game. It has markets, advanced industries, research centers, key manufacturers in specific segments, and the ability to shape global standards. But European advantage can no longer rest solely on the idea of a “trustworthy AI” or on the ethical superiority of its regulatory framework. Without enough homegrown infrastructure and companies capable of scaling like their competitors, the narrative risks sounding good in Brussels but falling short in the market.
The real question for the coming years is less ideological than it appears: will the EU translate its regulatory vision into a strong industrial and technological base, or will it continue to depend on those that have chosen to compete across the entire value chain? The AI Act gives Europe a voice. InvestAI and the AI Factories aim to build muscle. What remains uncertain is whether this will happen in time.
Frequently Asked Questions
What exactly is the European AI Act, and when does it fully come into force?
It’s the European law regulating AI based on risk levels. It took effect on August 1, 2024, and will be fully applicable—albeit with some exceptions and specific timelines—from August 2, 2026.
Is the EU only regulating AI, or is it also investing?
In addition to regulation, the Commission launched InvestAI to mobilize €200 billion, and it is promoting a network of 19 AI Factories and 13 AI Factory Antennas to expand computing capacity and infrastructure access.
Why is Europe still dependent on the US for technology?
Because a significant portion of cloud infrastructure, advanced models, core software, and key segments of the digital value chain are still dominated by US companies, while Europe has fewer global-scale companies in these areas.
Can Europe still compete with the US and China in AI?
Yes, but it must accelerate investments in infrastructure, talent, and enterprise scaling. The reports and plans from the European Commission acknowledge that regulation alone isn’t enough without simultaneously strengthening industrial and technological capabilities.
via: Portal Financiero