Control over key technologies and supply chains has become the strategic mantra of our time. But if national security logics keep crowding out economic ones, the promises of resilience will end up in economic loss. To reconcile security and economic competitiveness, asking the right questions is crucial: what role do economic interdependences still play in today’s global economy, and how can we expect them to evolve? The European Union (EU) has still to confront these questions. The global competition over technology––fueled above all by the transformative power of artificial intelligence (AI)––has spurred a frenzy of policy initiatives in the pursuit of the Union’s strategic autonomy and digital sovereignty. As of 2024, the EU imported 80% of its digital products, services, and infrastructure. The vast majority of chips used in the EU are designed in the United States (US) and manufactured in Korea, Taiwan, or China. The EU, alongside Japan, retains advantages in materials and equipment, but when it comes to AI more specifically, it is almost entirely dependent on American-designed hardware and cloud platforms.
Despite the future of technological relations with Beijing remaining unclear, Washington has been explicit: technological supremacy is a national security imperative, and the US must be the ruler of the emerging AI world. China, meanwhile, leveraging a maturing innovation ecosystem, can count on a vast domestic market and control over the global supply of critical raw materials. Additionally, China takes advantages in the prolongation of Russia’s war on Ukraine and the tests that it poses to the unity of the West. The EU sits uneasily amid great powers and their attempts to reconfigure the international system: still unprepared to take care of its own defence, it is fractured inside – and when it comes to digital technology, it is indispensable in certain niches but lacks the scale to compete.
Should Europe attempt to replicate global supply chains at home? Leaders have not yet answered this question. But the real problem is that this option would lie on a shaky assumption. In particular, that building digital sovereignty would be cheap and fast enough to make Europe withstand imminent external threats and internal fractures. Yet, building autonomy would require enormous economic resources – nearly €300 billion – that could conflict with overlapping challenges the EU faces, from military spending to ensuring climate resilience. Moreover, it will probably take too much time for Europe to replicate the equivalent of 80% of its digital imports. Estimates indicate nearly 10 years would be required. According to the European Court of Auditors, the EU is not even on pace to achieve the objective, included in the Chips Act, to produce 20% of global microchip by 2030.
Certainly, the intentions confirm that the EU leadership has acknowledged the complex environment it has to navigate and the entity of the techno-industrial calculations demanding a solution. To move ahead, however, more practical questions are to be asked. At present, the objective is to define a technological and industrial strategy grounded in the realities of an evolving global economy. This needs to start with a fairer assessment of global supply chains. Economic interdependences go a long way back and, despite recent attempts to reshape them, they keep supporting the global economy. Since at least the 1960s, national economies have relied on global supply chains to gain and maintain cost and quality advantages for their products. Companies enabled this process by proving the benefits of finding the most efficient solutions for the delivery of their products and services. Today, states are motivated to interfere with this process driven by national security concerns. Excessive intervention, however, complicates firms’ efficiency calculations and eventually, this could lead to higher costs of production, higher prices for end consumers, and lower prosperity overall.
These risks are particularly pertinent in the semiconductor industry. Relative to other industries, chips are very expensive products to make. They combine extreme capital and know-how intensity. One ASML lithography machine costs around $370 million; a state-of-the-art fabrication plant costs $20–40 billion. Today’s largest semiconductor manufacturer in the world, TSMC, spends nearly $40 billion on capital equipment and research every year. This is to say that in the course of the past decades, specialization has deepened so much that today, only a handful of firms possess the human capital that allows them to operate and profit at the cutting-edge. In turn, this makes leading firms less prone to substitution––an element that might resist the geopolitical re-engineering of supply chains.
Second, and relatedly, forward-looking strategies must get a sense of where markets are going. To understand the evolution of industries, the role of users and the international differences in the profile of demand are telling. Historically, major technological transformations––like those we are experiencing today––have challenged industry leaders, especially when accompanied by changes in the characteristics of market demand. To assess this in the context of chips and the future of the European industry, one should distinguish among different products, namely, mature and small-node chips. Mature-node chips (45 nm and above) are used primarily in the automotive and defence sectors as well as industrial applications. Small node chips (below 5-to-7 nm) are more energy efficient and faster, able to store more information, and thus lend themselves better to advanced technologies like AI, high-performance computing, datacenters, and smartphones.
In 2024, the global demand of chips by end use saw the biggest portion go to AI and computing (34.9%) and communications (33%), followed by automotive (12.7%), consumer electronics (9.9%) and industrial applications (8.4%). In Europe, by contrast, the vast majority of demand by domestic producers concerns the 65-90 nm nodes––namely chips that are 3-4 generations behind the frontier. Evidence is not conclusive on this matter, but it gives an indication of European prospects. The automotive industry, shifting to electric, will drive most local demand. However, while the growing sophistication of electric vehicles requires more sophisticated chips, that will most likely come from Asia, the outlooks of the European automotive industry are uncertain given the unresolved tradeoffs between decarbonization and competitiveness in Europe and the fierce competition coming from China. New opportunities may arise from surging defense expenditures, which could generate the scale and certainty investors require. But without comprehensive military and economic strategies, these demand-related uncertainties will possibly freeze Europe in a ‘middle-tech trap’, casting doubts over its capacity to compete at the frontier.
To be sure, acknowledging these facts is just a first step into the causes of European struggles and lack of strategy. In an age where the weaponisation of supply chains coexists with their continued, critical importance for the global economy and transformative technological evolutions, the path ahead will require the capacity of individuals, organisations, and states alike to navigate complex challenges with as much adaptation as vision.
Further Reading on E-International Relations
