The quest for self-sufficiency is as old as humanity. Tribes, city-states, and nations have long sought to bolster their independence. Now, amid growing geopolitical instability and supply chain uncertainty, the European Union (EU) is exploring ways to reduce its dependence on outside powers.
In June 2026, the European Commission announced a technological sovereignty package that included two legislative proposals—the Chips Act 2.0 and the Cloud and AI Development Act—aimed at boosting EU independence in semiconductors, cloud frameworks, and artificial intelligence.
“Right now, 90% of European digital infrastructure is not under the control of Europeans,” said Cristina Caffarra, an economist and co-founder of EuroStack, a nonprofit initiative that promotes EU tech sovereignty through strategic investment. “The question moving forward is whether Europe will be a sovereign continent or a digital colony.”
Despite political dimensions—tariffs, trade wars, and possible export controls top the list—the technical underpinnings and practical realities of a more independent EU are not trivial. Changes in the technology stack ripple across everything from electrical grids and hospitals to financial systems and cybersecurity. They also impact innovation, science, and research.
Escaping the orbits of the U.S., China, and Taiwan will not be easy. “The problem isn’t only how many billions, or tens of billions, are needed to build a leading-edge fab. It’s not having the skills and industrial infrastructure. This is a much more difficult game to win,” said Luca Benini, professor of electronics at Italy’s University of Bologna and chair of Digital Circuits and Systems at Switzerland’s ETH Zürich.
Power Struggles
Europe isn’t the only region coping with sovereignty issues. Every country is now at the mercy of others for materials, components, or products. The U.S. has focused on moving semiconductor manufacturing onshore. China, too, has spent hundreds of billions chasing chip and AI self-sufficiency—a task complicated by its inability to purchase the most advanced EUV lithography machines from Netherlands-based firm ASML.
What makes the EU different is its enormous size and power: 27 countries with a combined GDP of $19.5 trillion. Yet it remains heavily dependent on the U.S., China, and Taiwan for chips, cloud infrastructure, and AI. Even with some powerful players—ASML, STMicroelectronics, NXP, Infineon, Mistral, and software companies such as SAP and Dassault Systèmes—Europe lags on the global tech stage.
There’s a push to reshuffle the technology stack. “The historic ‘globalization’ of technology and supply chains has been a good move economically. However, when a supply chain is ‘stressed’ it can easily break,” said John Goodacre, professor of Computer Architectures at the University of Manchester in the U.K. “This stress, and other global tensions, has caused governments to reassess their dependencies . . . and what actions they should take to ensure a stress doesn’t lead to a collapse.”
This isn’t Europe’s first attempt at tech sovereignty. The European Chips Act 1.0, adopted in July 2023, aimed to double Europe’s global semiconductor market share to 20% by 2030. However, a 2025 European Court of Auditors (ECA) report found that the EU is unlikely to reach this goal: the Commission estimates that Europe will reach only 11.7% by 2030, up from 9.8% in 2022. Annemie Turtelboom, the ECA member who led the audit, described the 20% figure as “essentially aspirational,” noting it would require the EU to roughly quadruple production capacity.
A glaring example of the challenge is the Intel Magdeburg fab, a roughly €30 billion flagship investment and a centerpiece of Chips Act 1.0. Germany had committed nearly €10 billion in subsidies. Nevertheless, Intel formally cancelled the project—as well as a packaging plant in Poland—in July 2025. Anchor customers never materialized, and the project would have done little to boost sovereignty. Intel, after all, is a U.S. chipmaker.
The collapse of Magdeburg exposed a deep fault line. The EC has no mandate over industrial policy. Decisions about aid to specific companies rest with member states, said Tillman Schenk, a research assistant at the Bruegel economic think tank in Brussels. The result, he added, was “a very uncoordinated and fragmented policy. There was no grand vision of how this was supposed to go down.”
Schenk said Europe should stop chasing a fully self-sufficient technology stack and instead concentrate on the choke points where it already holds leverage. “If we cannot establish our own AI models and computing power in Europe in the near term, then let’s at least drive up the cost substantially for others to cut off access,” he said. “Let’s make ourselves so indispensable that such actions are very hard to do.”
Reinventing the Stack
Chips Act 2.0 would revise and update version 1.0. As part of the new plan, the European Commission hopes to build a sub-2nm leading-edge fab, better align EU semiconductor production with the strategic needs of industries and markets, and establish a more resilient supply chain, Schenk said. The plan also focuses on creating European buyers for European chips, something that version 1.0 conspicuously lacked. “They want to do everything, essentially,” he said.
Some leverage already exists. “The U.S. does not have a competitor to ASML,” said Hanna Dohmen, Senior Research Analyst at Georgetown University’s Center for Security and Emerging Technology. She added that the EU already has a web of sub-component suppliers, as well as Belgian semiconductor research institute imec. “It is important for Europe to recognize that kind of expertise and leverage in trade discussions with the U.S.”
To achieve greater sovereignty, the EU must increase investments in technology and build out a broad and strong ecosystem, Benini said. European fabless startups such as Axelera AI and SiPearl already are producing advanced AI chip designs. “Fabless companies have the potential to grow rapidly, and they don’t need a massive investment,” he said. A new wave of both government-sponsored and private equity funds—with golden-share provisions that block acquisitions by U.S. and Chinese mega-corporations—could unlock rapid development, he added.
Goodacre said that a reframing—adopting a focus on chiplets and components, rather than a leading-edge chip—would allow Europe to compete without having a multi-billion-euro production facility for specialized chips. EU projects like DARE and EPI, paired with a push toward RISC-V, an open instruction set architecture, could drive greater “accessibility and the ability to collaborate,” he said.
Another challenge is the time required to make changes, Goodacre said. For instance, Arm needed a decade—along with a foothold through cloud hyperscalers—before it gained widespread traction in datacenters. He doesn’t expect RISC-V to move faster. It will need credible chips that scale economically in a “nursery market.” The process could take 10 years, he noted.
Fabricating a Future?
Benini views the situation as a sustained transition rather than a big bang. Europeans must move past risk aversion and a highly fragmented approach to achieve an essential level of sovereignty. Yet, at the same time, he cautioned against a focus on total independence. “We live in a highly connected world and it’s unwise to build walls. A balanced approach is needed for hardware, software, AI, and other areas,” he said.
As Caffarra concluded: “Sovereignty means no more than that. We want a greater share of our own market. Not autarky, no protectionism, not decoupling.”
Samuel Greengard is an author and journalist based in West Linn, OR, USA.
Join the Discussion (0)
Become a Member or Sign In to Post a Comment