Add Silicon Canal to your Google News Feed. 
Europe is seeing more money flowing into artificial intelligence than ever before. Venture capital firms across the continent poured a record €10.4 billion into AI startups in 2024, according to Dealroom data. Government-backed funds in France, Germany, and the Netherlands have embarked on ambitious AI strategies. The European Investment Fund has allocated billions of dollars to deep technology. On paper, the ecosystem looks healthier than ever.
Despite this, the departure lounges at Schiphol, CDG, and Heathrow are always filled with one-way flights to San Francisco, New York, and Seattle. Europe’s best AI researchers and founders – the people this funding ultimately aims to support – continue to leave. The question is not whether there is a brain drain. That’s why record capital isn’t enough to reverse it.
canyon of compensation
Let’s start with the most obvious factor. Because it remains the most powerful factor. A senior machine learning engineer at a top European AI lab can expect to earn between €120,000 and €180,000 per year. Representatives from Google DeepMind’s U.S. office, OpenAI, or Anthropic typically earn a total compensation of $350,000 to $700,000, but can be more if equity is taken into account.
This stock element is very important. Amid the current AI boom, stocks in fast-growing U.S. AI companies could dwarf base salaries within a few years. When researchers see their colleagues who have moved stateside trading stocks at life-changing valuations, they instinctively feel the pull. It’s not greed, it’s rational behavior. Tax systems in Europe, particularly in countries such as Belgium, France and Germany, are compressing take-home pay even further, making inequality seem even wider.
The reason the record inflow of local money isn’t closing the gap is because money is flowing to different places. European venture capital firms raise money for companies rather than the level of individual compensation packages offered by American tech giants. A European startup that raised €20 million in Series A can’t compete with the talent acquisition budgets of companies that treat AI researchers like elite athletes.
Ecosystem impact: density creates gravity
Money alone cannot explain this pattern. There’s something more subtle at work, what organizational psychology researchers call the “agglomeration effect.” Clusters of top talent where other top talent is already in place. San Francisco’s AI ecosystem is now so dense in expertise, infrastructure, and ambition that it operates almost like a gravitational field.
When Demis Hassabis built DeepMind in London, it was a rare counter-magnet in Europe. But even DeepMind operates under the Alphabet umbrella and has considerable research capacity in the United States. Mistral AI in Paris and Aleph Alpha in Germany are genuine European candidates, but they remain exceptions rather than patterns.
Network effects are serious. In the Bay Area, researchers can have coffee with those building frontier models, have lunch with computing infrastructure leaders, and spend evenings discussing implementation challenges with policy experts. This kind of accidental conflict, the unstructured exchange of knowledge that fuels careers and ideas, is extremely difficult to reproduce when communities are scattered across a dozen European capitals with different languages, regulatory environments, and work cultures.
Regulation as a friction, not a barrier
It would be easy to blame EU AI laws and broader European regulatory instincts. But conversations with retired researchers suggest the situation is more nuanced. Most people do not cite regulation as the main reason for leaving. Instead, they describe it as friction. This is another factor that makes construction in Europe feel slower and more uncertain.
A more serious question is what the frictions say about institutional priorities. When the most visible AI policy outcome on the continent is a compliance framework rather than a computing infrastructure program, it sends a message about what matters. Researchers are carefully reading those signals. They don’t just choose a job. They are choosing the environment that evaluates their behavior.
What actually changes the equation?
The most commonly proposed solutions (more funding, more childcare facilities, more government AI strategies) address the symptoms rather than the root causes. Europe is not without money or ambition. There is a lack of structural conditions that make staying feel like the obvious choice.
Three shifts can truly move the needle. First is the competitive share structure and tax treatment of stock-based compensation. Some European countries still treat employee stock options more punitively than in the United States, making equity-focused packages far less effective at attracting and retaining top talent.
The second is intensive computing investment. Access to large GPU clusters remains a bottleneck for European researchers. The most ambitious research in AI requires infrastructure that few European institutions are currently able to provide. If you need 10,000 H100s for an experiment, go where the cluster is.
Third, and perhaps most importantly, there is a recalibration of the culture around ambition. Europe’s technology culture has improved dramatically over the past decade, but there remains a gap in what we might call “permission to think big.” The U.S. ecosystem encourages and even expects researchers to pursue transformative and paradigm-shifting research. European institutions often reward incremental progress and risk management.
The window is not closed, but it is narrowed
There are real reasons for optimism. France has been actively recruiting AI talent under President Macron’s administration, and has been achieving results. The Netherlands’ growing AI hub around Amsterdam and Delft continues to attract passionate researchers. The UK maintains a world-class AI research institution despite post-Brexit complexities.
However, the scope for Europe to build an autonomous AI talent ecosystem is shrinking. Every time a researcher leaves, the US cluster gets stronger and the European cluster gets weaker. Each departure increases the likelihood of the next departure. The feedback loop of tracking funding alone cannot be broken.
The continent that has produced some of the world’s best AI minds, from deep learning pioneers in European universities to the founders of companies now worth billions in Silicon Valley, needs to ask itself harder questions than, “How can we fund more AI?” We need to ask, “Why do the people we train continue to build their futures elsewhere?”
As unpleasant as it may be, the answer is not about money. It’s about creating an environment where the world’s most ambitious researchers feel like they can do their best research and get paid for it without hopping on a plane.
Featured image by Matheus Bertelli on Pexels
