Europe faces a critical dilemma: its landmark AI regulation may be undermining the very innovation it aims to protect. As the European Union enforces the AI Act, a growing chorus of technologists and policy experts warns that the compliance burden is scaring away investors and driving European AI talent across the Atlantic, leaving the continent further behind in the global race for technological leadership.

Is the EU AI Act Backfiring on European Startups?

The concern centers on how the AI Act's classification system works. As AI models grow more capable, they risk crossing regulatory thresholds that trigger "systemic" classification, subjecting them to onerous compliance requirements. For European AI entrepreneurs, this creates a chilling effect before their companies even scale. The threat of sudden regulatory burden discourages investors from funding European AI ventures, while simultaneously encouraging founders to relocate their operations to less restrictive jurisdictions in North America.

"Europe should instead repeal the EU AI Act, which has chilled investment in the European AI sector. European AI entrepreneurs like me worry that their models will soon reach the AI Act's thresholds, causing them to be classified as 'systemic' and subject to onerous compliance requirements," said Garvan Walshe.

Garvan Walshe, Former Foreign Policy Adviser to the British Conservative Party and Co-founder of Kronkite

This regulatory squeeze arrives at a particularly vulnerable moment. Europe already lags significantly behind the United States and China in scaling large, influential AI companies. The continent holds a distant third-place position in the global AI race, coupled with a limited track record of commercializing innovations and retaining tech talent. Over 85% of European cloud infrastructure, which serves as the backbone for AI development, is controlled by US-based providers as of 2025. This dependency on external infrastructure compounds the challenge of building homegrown AI capabilities.

What Are Europe's Structural Barriers to AI Leadership?

Europe's technology landscape reveals significant regional strengths that remain underutilized. Germany and Austria lead in business-to-business software and industrial automation; France excels in AI research and aerospace technology; Nordic countries specialize in digital finance and clean technology; and the Netherlands and Belgium focus on logistics software and semiconductors. Yet these pockets of excellence have not translated into the kind of large-scale AI companies that dominate global markets.

The broader challenge extends beyond regulation. Europe's geopolitical vulnerabilities expose a critical dependency: the continent relies heavily on external sources for essential materials and services. Recent disruptions have highlighted these risks to both national and regional resilience. Additionally, the AI sector itself faces a fundamental profitability question that affects investment globally. Large language models, the foundation of most AI applications, are trained on similar datasets and detect comparable patterns, making them largely undifferentiated commodities despite trillions in investment.

The economics of AI development reveal another troubling pattern. AI model developers face intense pressure to obtain critical semiconductor supplies, spending the bulk of their capital on chips rather than differentiation. Chipmakers have become so profitable that they are funding their own customers to purchase their products, capturing the majority of value in the AI supply chain. When European policymakers subsidize AI gigafactories through public funding, they risk making the same mistake as private investors: pouring money into a sector that may not generate the returns needed to satisfy investors or justify public expenditure.

How Can Europe Recalibrate Its AI Strategy?

• Reconsider Regulatory Thresholds: Policymakers should evaluate whether the AI Act's systemic classification triggers are calibrated appropriately for early-stage companies, or if they inadvertently penalize European startups before they achieve scale.
• Invest in Next-Generation Research: Rather than subsidizing gigafactories to produce copycat large language models, public funding should target fundamental research into emerging AI technologies and capabilities that don't yet exist.
• Leverage Regional Strengths: Europe should build on its existing advantages in business-to-business software, industrial applications, specialized small and medium-sized enterprises, and high-quality standards rather than attempting to replicate Silicon Valley's model.
• Harmonize Cross-Border Rules: Europe faces over 20 different national regulatory regimes that need alignment, unlike the more unified approaches in the US and China, creating compliance complexity that smaller companies cannot easily navigate.

Walshe's perspective reflects a growing frustration among European tech founders who see the regulatory environment as a barrier rather than a safeguard. The AI Act was designed to protect citizens and ensure responsible development, but its implementation may be having the opposite effect on the continent's ability to compete globally. The challenge is not whether regulation is necessary, but whether Europe's current approach strikes the right balance between safety and innovation.

Europe's path forward requires acknowledging a hard truth: the continent cannot win a race to build the largest or cheapest large language models. Instead, success depends on cultivating innovation rooted in Europe's unique strengths and values. This means supporting deep expertise in specialized domains, fostering cross-border cooperation within the single market, and creating regulatory environments that encourage experimentation rather than discourage it. The next decade will determine whether Europe becomes a strong, resilient technology hub or continues its slide toward dependency on external innovation.