EU: 80% of AI from Abroad—Aim to Build European Language Models

For the AI industry, the EU’s 80% figure crystallizes a strategic vulnerability: European enterprises and governments are overwhelmingly dependent on US-built models like GPT and Gemini, trained on English-centric data. The bloc’s new plan to invest in AI trained in European languages and content, combined with the Brazil digital deal, signals a future where AI development shifts toward more culturally diverse and regulation-compliant datasets. This could fragment the global AI model market but also create niche opportunities for European and Latin American AI startups.

The European Union's technology sovereignty chief, Henna Virkkunen, issued a stark warning at the Web Summit Rio on June 12, 2026, regarding the bloc's heavy reliance on US technology companies in critical sectors such as cybersecurity and defense. Her remarks, delivered to journalists on the sidelines of the conference, emphasized that 80% of Europe's technologies, including artificial intelligence, originate from outside the continent. Virkkunen made clear that the EU's pursuit of digital sovereignty is not protectionist but a strategic imperative to avoid over-dependence on third countries, particularly the United States. This statement comes just weeks after the EU unveiled a comprehensive plan to reduce its technological dependency on both American and Asian firms, signaling a pivotal shift in the bloc's industrial and regulatory posture.

“

The European Union's technology sovereignty chief, Henna Virkkunen, issued a stark warning at the Web Summit Rio on June 12, 2026, regarding the bloc's heavy reliance on US technology companies in critical sectors such as cybersecurity and defense.

The EU's digital sovereignty agenda has been building for years, fueled by concerns over data privacy, supply chain resilience, and the dominance of US cloud providers, semiconductor firms, and AI developers. The bloc's regulatory arsenal—including the General Data Protection Regulation (GDPR), Digital Markets Act, and the upcoming AI Act—has already imposed significant compliance costs on non-European companies. Virkkunen's comments now explicitly link these regulatory efforts to a broader geostrategic drive to cultivate homegrown technological capabilities. The 80% dependency figure underscores the scale of the challenge: European businesses and governments rely heavily on Amazon Web Services, Microsoft Azure, Google Cloud, and AI models from OpenAI and Meta, creating vulnerabilities in an era of geopolitical fragmentation.

From an industry perspective, the EU's push could reshape market dynamics across multiple sectors. For US tech giants, the immediate risk is not just regulatory friction but the gradual erosion of market share as European alternatives gain state-backed momentum and preference in public procurement. Companies like SAP, ASML, and emerging AI startups such as Mistral AI and Aleph Alpha could benefit from increased EU funding and policy tailwinds. The EU's plan to invest in training AI models in European languages and content aligns with its cultural sovereignty goals, but it also requires massive investment in compute infrastructure—a domain where US-based NVIDIA currently dominates. This creates a delicate balancing act for European policymakers who must attract foreign investment while building local capacity.

The announcement of Brazil as the EU's fifth digital partner—following Japan, Canada, Singapore, and South Korea—highlights the bloc's strategy of forging alliances with like-minded democratic nations to create interoperable digital standards. Brazil, with 160 million internet users and a top-10 technology market, offers a significant opportunity for European companies to expand their digital services and for joint AI governance frameworks. The partnership deal, to be signed on June 13, 2026, covers data governance, AI, digital infrastructure, and platform regulation, potentially creating a template for future agreements with other emerging economies. This also positions the EU as a normative power, exporting its regulatory philosophy through bilateral ties.

The implications for the startup ecosystem are profound. European venture-backed tech firms, particularly in AI and cybersecurity, stand to gain from preferential access to EU research grants, procurement contracts, and a regulatory environment that favors local champions. However, the shift also introduces uncertainty: startups that rely on US cloud platforms or seek US market expansion may face dual compliance burdens. Moreover, the EU's emphasis on 'values-aligned' technology could fragment the global digital economy into geopolitical blocs, complicating cross-border data flows and scaling strategies.