There are basically only two competitors in artificial intelligence: the United States and China. Europe is nowhere to be found.
This is not a rhetorical flourish. It is a measurable fact. EU member states account for just 5 percent of global AI computing capacity, while the United States controls approximately 75 percent and China roughly 15 percent. Of the notable AI models developed last year, the United States produced 59 and China 35. France and the United Kingdom each produced one. Moreover, the gap is not narrowing; it is widening.
And yet Europe has the scientists. It has the universities. It has the industrial know-how—the machine tools, the precision engineering, the decades of manufacturing data that no Silicon Valley firm can replicate. The question, then, is not why Europe lacks capability. It is why that capability fails to translate into AI power. The answer lies in structures of interest, military logic, and monetary architecture that emerged long before the current AI race began.
Rich interest groups want America to dominate
American interest groups want the United States to dominate AI. This should surprise no one. What is less obvious, however, is how deeply those groups are interconnected with European counterparts—and how that interconnection serves American dominance rather than European autonomy.
The mechanism is straightforward. American rich families, tech firms, venture capital funds, and defense contractors operate within an ecosystem that rewards concentration. In 2025, U.S.-based AI companies attracted about $194 billion in venture capital—roughly 75 percent of global AI VC investment. EU companies received about $15.8 billion, or 6 percent. Moreover, mega-deals worth more than $100 million accounted for about 73 percent of all AI venture investment worldwide.
European capital exists. German companies and insurers hold substantial reserves that they could mobilize for AI infrastructure. However, European finance is conservative, oriented toward traditional banking and short-term returns. As a result, a structural asymmetry emerges: a French AI founder like Arthur Mensch of Mistral raises €3 billion, while his smallest American competitor raises €30 billion. The interest groups that could fund European AI instead channel capital into American platforms and American returns. At the same time, European institutional investors, pension funds, and banks are deeply integrated into dollar-denominated markets. Their incentives align with the American ecosystem, not against it.
This is not a conspiracy. Rather, it is a coordination problem dressed as a capital allocation problem. European elites benefit from the existing arrangement. American dominance in AI is not simply imposed on Europe; European capital seeking American returns helps finance it.
AI = military research, US wants to be first
Silicon Valley is not simply a technology hub. It is a military-industrial complex in its current form, and AI is its central weapon.
The pattern is well documented. In 2025, the Pentagon awarded contracts worth $200 million to OpenAI, xAI, Google, and Anthropic for “frontier AI” projects. The U.S. Army awarded Anduril—a defense technology company founded by a former Palantir executive—a $20 billion contract for AI-backed software to run on military systems. Meanwhile, Palantir’s Project Maven, which integrates Anthropic’s AI with Palantir’s data systems, was used in the 2026 Iran war to process targeting data at a scale of 5,000 targets per day, compressing kill-decision time to 72 seconds.
This is the “military-industrial complex” that Eisenhower warned about, reengineered for the digital age. Unlike the Cold War version, this formation is faster, transnational, and vertically integrated—extending from cloud platforms and AI models to autonomous drones and military satellites. Pentagon officials have also begun commissioning Silicon Valley executives directly into military ranks. Meanwhile, the Defense Advanced Research Projects Agency (DARPA), the Defense Innovation Unit (DIU), and In-Q-Tel, the CIA’s venture capital arm, have institutionalized the flow of private technology into military applications.
Europe has no equivalent. No European defense ministry has an AI budget on this scale, and European countries lack domestic frontier AI companies that could contract with their militaries even if they wanted to. Germany’s Helsing, the most prominent European defense AI startup, has raised €450 million—a fraction of the American contracts awarded to a single firm like Anduril. Consequently, European militaries that want AI capability must buy American. The military logic that drove the original internet, GPS, and semiconductor industries is now driving AI, and Europe is a customer, not a producer.
Dollar and economic dominance and educational dominance
The final reason is the deepest and the least discussed: the dollar system and the educational pipeline that sustains American AI dominance are two sides of the same coin.
The dollar’s role as the world’s reserve currency is not a neutral fact. Around 60 percent of global reserves are held in dollars, and over 80 percent of trade finance is settled through it. This gives the United States an extraordinary capacity to mobilize capital. When the Federal Reserve raises interest rates, weaker economies are pushed into crisis. When the U.S. government runs deficits, the world absorbs dollar-denominated assets. In turn, this financial supremacy provides the material basis for American research; without it, U.S. science would look fundamentally different.
AI is now being deliberately bound to this monetary architecture. The United States is mobilizing capital on an unprecedented scale—$800 billion in AI capital expenditure in a single year, financed through private equity, public debt, and stock markets. As Deutsche Bank has observed, the dollar is becoming directly leveraged to the AI race. In other words, a new “technodollar” system is emerging, underwritten by digital infrastructure and the global scramble for AI capacity.
Brain drain
The educational pipeline reinforces this dynamic. U.S. universities function as a “legal ladder” into the American AI sector, offering lawful entry through F-1 visas and embedding students inside the country’s leading research-industry networks. Seventy percent of full-time graduate students in AI-related fields in the United States are international students, and 70 percent of immigrant founders of the most promising AI firms came to the United States on a student visa. As a result, the United States has attracted and retained twice as much top AI talent as it has supplied into the market, drawing disproportionately from Europe, Israel, and Canada.
Europe trains the talent. The United States employs it. Approximately 40 percent of AI doctoral graduates from European universities emigrate to the United States within five years of graduation. For graduates of Europe’s top computer science programs, the figure approaches 30 percent. This is not a failure of European education. Instead, it is a success of the American system—a system that combines dollar-denominated salaries, frontier research infrastructure, and immigration pathways that European countries have not built.
Conclusion
Europe is not lagging in AI because it lacks scientists, universities, or industrial knowledge. It is lagging because the structures that govern AI—capital flows, military procurement, monetary architecture, and talent pipelines—were designed to produce American dominance. European interest groups are integrated into those structures. European militaries depend on American AI. Meanwhile, Europe trains talent that the United States ultimately employs.
This is not a problem that a new funding program or regulatory simplification can solve. It is a problem of architecture. Until Europe builds its own capital markets, its own defense AI ecosystem, and its own pathways for retaining the talent it trains, it will remain what it is today: a supplier of inputs to an American AI system, and a consumer of its outputs.
Leave a Reply