Quick Buzz Feed
In its bid to compete with the US on AI, Europe could learn from both China and from the classic Airbus industrial policy case. The global race for artificial intelligence leadership is largely a contest for advanced computing power, with the US dominating chip design and China aggressively building its domestic capacity. Europe, despite possessing key upstream hardware assets, risks losing economic autonomy if it remains dependent on foreign AI compute infrastructure. This article outlines a strategic approach for Europe, drawing parallels with China's two-track strategy and the successful Airbus industrial policy model, to foster its own competitive AI hardware industry and secure its future in the digital economy.
China is aggressively working to close its AI hardware gap with the United States through a comprehensive two-pronged strategy. Diplomatically, it leverages its significant economic influence and control over critical resources, such as rare earths, to negotiate concessions on US export controls, leading to some easing of restrictions on advanced AI hardware. Domestically, China is fostering a robust, self-sufficient AI chip ecosystem. Key players like Huawei, with its Ascend chips, and other domestic firms such as Alibaba and Baidu, are capturing an increasing share of the Chinese AI chip market. This domestic growth is strongly supported by government directives that prioritize local hardware for state-owned enterprises, creating a captive market. This strategic approach ensures a steady revenue stream for domestic companies, which is then reinvested into R&D, fueling continuous improvements in chip capabilities and offering valuable lessons for Europe in achieving strategic self-reliance in this critical technology sector.
While current Chinese AI chips, such as Huawei's Ascend 910C, may still lag behind NVIDIA's top-tier offerings in raw performance and the maturity of its software ecosystem (like CUDA), China is strategically addressing these disparities. Huawei has ambitious plans to significantly enhance its chip capabilities, aiming to double them by the end of 2027. China's broader strategy is to use short-term reliance on advanced US chips as a stopgap while simultaneously driving the widespread adoption and continuous improvement of its indigenous chips. This is facilitated through various support mechanisms, including subsidies, guaranteed domestic demand, and offering favorable energy prices to AI firms that utilize local chips. This indicates that the existing performance gap and the challenges of migrating from established software platforms are perceived as transitional, and are expected to diminish as Huawei’s technological stack and its CANN platform mature over the next three to five years, thereby solidifying China's long-term strategic autonomy in AI compute.
Europe faces significant challenges in establishing a competitive AI presence, primarily due to a critical deficit in the scale of compute resources and capital investment. European AI initiatives, exemplified by France's Mistral, frequently depend on external infrastructure, such as Microsoft Azure's supercomputing capabilities, and rely on Silicon Valley venture capital to attract talent and funding. Unlike China's unified approach, the European Union currently lacks a coordinated mechanism to direct public procurement towards bolstering its domestic AI hardware R&D. This results in fragmented demand across its twenty-seven member states, preventing the creation of a large enough 'captive market' that could generate the substantial revenues necessary for continuous innovation. Although there are emerging plans, such as a 'European preference' in public procurement for critical technologies and sovereign cloud initiatives, these efforts are still in their nascent stages and have limited leverage over private-sector demand, making it difficult for Europe to replicate the successful domestic market dynamics seen in China.
Europe can draw critical insights from the successful Airbus industrial policy model to formulate its AI hardware strategy. The core principle for an effective strategy should be the orchestration of genuine technological complementarities and comparative advantages among European firms, rather than politically driven allocations of work across member states. Examples include ASML's global dominance in extreme ultraviolet lithography, IMEC's leadership in advanced semiconductor research, and the expertise of companies like Infineon and STMicroelectronics in power semiconductors and Carl Zeiss in precision optics. The institutional structure must evolve beyond a loose consortium to an integrated entity with unified management, shared intellectual property, and the autonomy to make market-driven investment decisions. Existing EU frameworks, such as Important Projects of Common European Interest (IPCEIs) and the European Chips Act, offer legal and financial tools to facilitate such joint R&D and industrial expansion. While directly replicating China's coercive demand-side strategies is not feasible, coordinated public procurement for AI compute in critical public sectors (administration, healthcare, defense, research) could create a substantial initial demand base. This approach, however, necessitates an honest reckoning with potential short-to-medium-term performance penalties for European companies choosing domestic alternatives, implying the need for compensation mechanisms (e.g., R&D subsidies, preferential pricing for public contracts) to incentivize adoption and ensure long-term strategic autonomy. Such an initiative demands a long-term commitment, akin to the two decades it took for Airbus to become genuinely competitive.