- China’s AI Talent Camps – and What Can Europe Do? - 29 April, 2026
- (Analysis) When Taiwan Becomes the Flashpoint: Inside the Latest Japan-China Tensions - 20 April, 2026
- China’s AI Talent Camps – and What Can Europe Do? - 29 April, 2026
- [REPORT] Greenland on the Chessboard: European Strategic Autonomy Challenge - 22 April, 2026
Introduction
In less than a decade, the global AI race has led to a competition between countries, among which the United States, China, and Europe; all trying to get the biggest share in the market. In this race, people matter just as much as chips and capital. While Silicon Valley has set the pace through Big Tech-labs, China has been cultivating something far less visible: gigantic talent pools for math and informatics talents that form the backbone of its AI sector. Europe, despite housing world-leading universities, remains largely a regulatory power and a spectator watching the US and China dictate the rules of the “AI game”. The time has come for Europe to decide how to shape its own AI talent and emerge as a serious competitor.
Chinese “Genius Classes”
China’s AI breakthrough didn’t come out of nowhere: it was the product of a strong, organized talent pipeline spanning from the 80s. In an attempt to capitalize on AI’s potential, China has developed a program in which roughly 100,000 teenagers, from the age of 16, are selected each year for so-called “genius classes” at elite schools, where they are trained for international Olympiads in mathematics, physics, informatics, and related sciences. These students are temporarily exempted from their standard school curriculum, and, over two years, work through the full secondary syllabus alongside portions of undergraduate content. This academic intensity typically comes at the expense of subjects such as history and politics.
University Shortcut
One of the reasons so many teenagers are interested in this program is that whoever gets to the top of these national competitions is directly admitted into China’s top universities, such as Tsinghua and Beijing University. Besides this, other selective pathways have also emerged, such as the “Yao Class” in computer science at Tsinghua, which annually admits only a few dozen of the very best Olympiad winners. The result is a scalable, competition-driven infrastructure that systematically enables “mass production” of AI talent. The names that come forward from these programs show how impactful they actually are: the founder of ByteDance, leading figures of Taobao and PDD, the billionaire founder of Meituan, the engineers behind DeepSeek’s R1 model, and the brothers who co-founded the chip designer Cambricon all passed through these genius classes.
China’s Main Driving Force in Industrial Upgrading
China’s genius classes strongly resemble honors programs or Olympiad programs found in the West, but their meaning extends far beyond mere individual prestige and accomplishment. The programs have been embedded into China’s New Generational Artificial Intelligence Development Plan (AIDP), which since 2017 has formally positioned AI as the primary driving force behind the country’s industrial upgrading and economic transformation. Olympiads are not only a side issue or part of education culture; they are perceived as a geopolitical instrument generating soft power.
Generating Soft Power
China shows that it can compete with, and eventually surpass, traditional scientific powers such as the US and Europe. The international success of Chinese students attracts both foreign undergraduates and multinational companies seeking access to a large pool of highly skilled engineers. Beyond visibility, these elites circulate through global firms, ensuring the growth of the Chinese industrial ecosystem [i1] and extending its perspectives on AI applications and governance. Domestically, this system also acts as a powerful source of legitimacy for the party-state. The narrative that the Communist Party has cultivated a generation of geniuses through central planning, thereby enabling the nation’s technological success, supports the claim that authoritarian rule delivers efficiency and long-term vision, in contrast to liberal democracies often portrayed as slowed by deliberation and internal debate. In the context of the global AI race, China’s talent strategy is therefore not only a human capital policy, but also a form of ideological and technological soft power that shifts the balance of power with the US and exerts pressure on Europe’s technological capabilities.
Europe’s Response
The global AI race is increasingly a contest over talent formation, knowledge control, and the ability to shape the rules of technological development. China understood this earlier than most Western countries: its long-term investment in elite STEM cultivation created a pipeline feeding top universities, strategic industries and advanced AI research, effectively turning education policy into state capacity. In Europe, that shift is visible even within universities. ETH Zurich has formalised security screening for certain master’s, doctoral, employment, and guest applications linked to sanctioned or proliferation-risk countries, reflecting the growing view that advanced STEM education is no longer just an academic matter but one tied to economic and national security.
At the regulatory level, the EU has responded with the AI Act, which entered into force on 1 August 2024 and began phased application in February 2025. Yet, the Act is more than a compliance framework; it is also an expression of strategic intent. By promoting a human-centric and trustworthy AI model, Europe is seeking to project regulatory soft power internationally while reducing dependence on both US platform dominance and Chinese technostate expansion. The problem is that soft power alone does not create secure frontier leadership. It can shape norms and market behaviour, but without industrial scale, talent concentration, and technological depth, it cannot by itself deliver strategic autonomy.
Is It Enough?
Europe’s response, then, is not absent. The real question is whether it is moving with sufficient speed, coherence, and scale. On research funding, Horizon Europe remains the Union’s central instrument, with an indicative budget of €93.5 billion for 2021 to 2027, while the Commission reports that €2.6 billion was allocated to AI research and development in 2021 and 2022 alone. (European Commission, Directorate-General for Research and Innovation). More recently, the European AI in Science Strategy has signaled a more serious effort to organise Europe’s fragmented strengths by linking talent, computing infrastructure, and funding, financing for networks of excellence and doctoral training. While the wider strategy aims to expand access to computational power and raise annual EU AI investment to more than €3 billion. On an industrial scale, the AI Content Agenda and InvestAI push beyond research toward capability-building, seeking to mobilise support such as investments for up to five AI gigafactories, while 19 AI factories are intended to support startups, industry, and researchers across Europe. The Chips Act follows the same strategic logic: strengthening the semiconductor ecosystem, increasing supply-chain resilience, and reducing external dependencies. Taken together, these measures show that Europe is no longer relying on regulation alone. It is trying to combine industrial policy, computing infrastructure, research funding, and rule-setting into a more coherent AI strategy. Its weakness is not diagnosis, it is execution.
Conclusion
China’s AI dominance is the result of decades of state-led investment in scalable talent pipelines. By systematically selecting and training hundreds of thousands of teenagers through genius classes, it has built a form of soft power that is reshaping the global AI landscape. Partly in response to China’s rapid acceleration, Europe has finally recognized that AI is a sovereignty issue, not just a digital one. As a result, the EU is funding research, building computing capacity, strengthening semiconductor resilience, and using regulation as both a governance tool and a form of strategic soft power. But it still needs more: earlier STEM identification, stronger elite training pipelines, better researcher retention, deeper scale-up financing, and much faster translation of university research into industrial leadership. Europe’s real opportunity is to fuse hard capacity with normative influence. If it succeeds, it will not only regulate the AI age, it will help define it.
Questions
- Can Europe ever build a talent pipeline that rivals China’s in scale and ambition, or is its commitment to academic freedom and open competition structurally incompatible with the kind of state-led cultivation that produced DeepSeek and ByteDance?
- Is Europe’s regulatory approach a genuine industrial strategy, or an attempt to shape a race it has already fallen behind in?
- In an era defined by technological sovereignty, can Europe afford to remain a normative power in AI while others build the frontier?
- Does Europe need civil-military AI integration to achieve sovereignty?
- Are ethical AI frameworks meaningful without industrial leverage?
- Should Europe prioritize a few AI champions rather than a competitive market?
Further Reading
