Europe Faces Critical Choice in Global AI Race as Energy Costs and Geopolitical Dependencies Collide

The Open Source Dilemma

Eric Schmidt issued a stark warning on Tuesday that Europe risks becoming dependent on Chinese artificial intelligence models unless it makes substantial investments in its own open-source AI capabilities. Speaking at the Swiss gathering, Schmidt outlined how the global AI landscape is splitting along distinct lines.

In the United States, companies are largely moving toward closed-source models that require licensing and payment. China, by contrast, is pursuing an open-weight, open-source approach that makes its models freely available. Schmidt argued that without significant European investment in homegrown models, Europe will inevitably default to Chinese AI systems, not because they are superior, but simply because they are free.

This trajectory concerns Schmidt because it creates profound geopolitical dependencies. European companies, governments, and institutions could find themselves reliant on AI infrastructure controlled by a geopolitical rival, with all the strategic vulnerabilities that entails. Chinese models from companies including DeepSeek and Alibaba have already gained substantial global traction, with some United States technology firms incorporating them into applications.

Energy as the New Competitive Frontier

Satya Nadella framed the AI competition through an entirely different lens during a separate session at Davos, identifying energy costs as the decisive factor that will determine which nations dominate the artificial intelligence economy. Nadella introduced the concept of AI tokens as a new global commodity, representing the basic units of processing that power AI tasks.

GDP growth in any location will be directly linked to the energy costs associated with utilizing AI, according to Nadella. This creates a significant challenge for Europe, which has struggled with some of the highest energy prices globally following sanctions related to Russia's invasion of Ukraine in 2022. The economics are stark. If producing AI tokens costs substantially more in Europe than in competing regions, European firms cannot compete effectively in AI-driven markets.

The scale of the energy challenge is considerable. According to OECD research, the European Union's electricity-intensive industries, including the data centres essential for AI operations, face persistent cost disadvantages. Demand from data centres is projected to surge from 96 terawatt-hours in 2024 to approximately 236 terawatt-hours by 2035. This represents more than a doubling of electricity consumption for AI infrastructure alone.

Converging Pressures on European Competitiveness

Both warnings point toward the same fundamental outcome, Europe's diminishing ability to participate as a major player in the AI-driven global economy. Schmidt's concern centres on technological sovereignty and avoiding dependency on Chinese AI systems. Nadella's focus is on the raw economics of AI production and the reality that high energy costs could price European companies out of competitive AI development entirely.

The forum's record government participation, with more than 60 heads of state and government attending alongside central bank governors and ministers from over 130 countries, underscores how AI has transcended its origins as a technology industry concern to become a central question of national competitiveness and economic security.

Europe's Strategic Responses

The European Commission has begun addressing these challenges, recently launching consultations on strengthening the continent's open-source AI ecosystem. The commission has acknowledged that much of the value generated by open-source projects is currently exploited outside the European Union, a recognition that aligns with Schmidt's warnings about dependency on external AI systems.

However, the dual nature of the challenge facing Europe makes solutions complex. Investing heavily in domestic AI development addresses the sovereignty concerns Schmidt raises, but it does not resolve the energy cost disadvantages that Nadella identifies as fundamental to AI competitiveness. Building European AI capabilities while operating with substantially higher electricity costs than competitors creates a structural economic disadvantage.

The Global AI Race Context

The Atlantic Council has noted that the AI race in 2026 remains defined by a multipolar order in which the United States and China will continue to yield the greatest influence. Middle powers, including European nations and India, are working to close the gap but face substantial challenges in marshalling the resources, both financial and energetic, required to compete at the frontier of AI development.

China's approach to AI governance is evolving from principle-based guidance toward more institutionalized and systematic frameworks, positioning Beijing as a contributor to global rule formation rather than merely a participant. This strategic positioning, combined with China's commitment to open-source AI models, creates appeal for nations and companies seeking accessible AI capabilities without the high costs associated with proprietary Western systems.

Implications for Global AI Development

The contrasts emerging at Davos 2026 reveal how artificial intelligence has become inseparable from questions of energy security, economic competitiveness, and geopolitical positioning. Nadella emphasized that society will quickly lose the social permission to use energy for generating AI tokens if those tokens are not demonstrably improving outcomes in healthcare, education, public sector efficiency, and private sector competitiveness.

This places additional pressure on European AI development to not only overcome cost disadvantages and avoid foreign dependencies but also to demonstrate tangible societal value that justifies the substantial energy consumption required. The intersection of these challenges represents perhaps the most complex strategic puzzle facing European policymakers as they attempt to chart a viable course in the global AI race.

The warnings from Schmidt and Nadella at Davos make clear that the window for Europe to establish itself as a competitive force in artificial intelligence may be closing rapidly. Without substantial action on both technological investment and energy infrastructure, Europe risks finding itself relegated to a secondary position in what many see as the defining technological competition of the 21st century.