Joe Tsai, Chairman of Alibaba Group Holding Ltd., delivered a blunt message to European leaders at the Wave by Vento conference in Turin on October 7, 2026: technological dependence is a vulnerability Europe can no longer afford. He argued that open-source artificial intelligence offers Europe its most viable route to genuine technological autonomy from both the United States and China.
The warning speaks to a strategic anxiety that’s been growing across European capitals for years. Europe doesn’t have a Microsoft, a Meta or a Google. In the race for AI dominance, the continent risks relying on technology stacks controlled by American or Chinese corporations, undermining both its competitiveness and its sovereignty.
The Sovereignty Dimension
Tsai put the stakes in terms European politicians can’t ignore. “No country would want to entirely trust another country's technology: if there's a change in government, a change in circumstances, they can just shut off the technology from you,” he said. His warning reflects a reality already driving policy debates in Brussels: technological dependence creates political vulnerability. If Europe runs its AI infrastructure on closed-source American systems, American policy decisions—regulatory shifts, export controls or geopolitical disputes—become Europe’s problem.
This isn’t abstract theorizing. The US has already weaponized technology controls against China, while Europe has watched American firms navigate a labyrinth of export restrictions. Tsai’s message is direct: waiting for American companies to share their models, or hoping Chinese competitors will compete fairly, is naive. Europe needs its own computing stack.
Building European Capacity
Tsai’s prescription is concrete: Europe must invest in computing infrastructure built to train and run AI models. He isn’t calling for protectionism or anti-American sentiment. It’s a pragmatic case for strategic autonomy. Open-source AI, with models and tools available to researchers and companies across Europe, would spread capability widely instead of concentrating it in a handful of Silicon Valley headquarters or Beijing boardrooms.
The practical challenge is substantial. Competing at scale requires not just capital, but also talent, energy security and regulatory frameworks that let infrastructure function. European governments and companies would also need to coordinate at a level that’s historically been difficult to achieve.
The alternative Tsai describes is grimmer. Technological lock-in deepens over time. Once European companies and research institutions depend on American or Chinese AI platforms, switching costs become prohibitive. Sovereignty, in this framing, is something you have to build before you lose it.
Why This Matters:
Tsai’s argument aligns with a core center-right concern about European autonomy within a reformed EU framework: no member state—and no union of member states—should depend structurally on the strategic choices of external powers. The question of AI infrastructure isn’t primarily about innovation prizes or venture capital. It’s whether European governments can make consequential technological and economic decisions without asking permission from Washington or Beijing. Building European open-source AI capacity would require coordination between member states and private industry, but it would preserve the possibility of genuine strategic choice. Continued technological dependence steadily erodes the practical meaning of European sovereignty.