🇪🇺 EU AI & Sovereignty — June 23, 2026

1. Anthropic shutdown ignites calls for sovereign AI across Europe — The US blocked Anthropic’s Fable 5 and Mythos 5 models for non-American users, sending shockwaves through Europe. Mistral AI CEO Arthur Mensch called it an “existential wake-up call.” The G7 summit in June 2026 centered on European AI autonomy as a direct consequence.
   Source: Fortune / Euronews / The Verge
2. EU Cloud and AI Development Act (CADA) published — The European Commission presented a legislative framework to reduce dependency on non-EU cloud and AI services. Reception is mixed — praised for ambition but criticized for fragile foundations and implementation risks.
   Source: European Commission / Taylor Wessing / Hogan Lovells
3. Mistral AI targets €20B valuation with €3B funding round — Mistral AI is in talks for a €3B investment round valuing the company at €20B. CEO Arthur Mensch warns Europe has 2 years to avoid becoming a US “vassal state” on AI. The company is building an Nvidia-powered AI hub in Paris.
   Source: TechCrunch / Startup Fortune / Business Insider
4. EU AI Act deadline delay: agreement on high-risk system postponement — EU lawmakers reached a provisional agreement to delay the deadline for high-risk AI system obligations. Critics warn this slows oversight of dangerous applications. New bans (including AI nudifiers) are being accelerated.
   Source: Sidley Austin / S&P Global / Stibbe
5. France and Germany call for European AI sovereignty at VivaTech 2026 — At VivaTech 2026 in Paris, French and German government representatives called for a joint European AI strategy. The event marked the transition from “planning” to “building phase” for European AI infrastructure. A Franco-German AI center was announced.
   Source: Euronews / Tech Times
6. EU Tech Sovereignty Package: Commission aims to reduce foreign tech dependency — The EC’s Tech Sovereignty Package includes measures for digital independence, with a surprisingly large role for open source. Critics speak of “sovereignty washing” and question whether the plans have sufficient teeth. Open-source hardware is named as a crucial prerequisite.
   Source: European Commission / Global Policy Watch / Tech Policy Press
7. UK launches £500M Sovereign AI Fund for independent AI infrastructure — The United Kingdom launched a £500M (€675M) Sovereign AI venture fund aimed at reducing foreign tech dependency. The fund invests in 7 companies and aligns with a broader European trend of state capital for AI sovereignty.
   Source: Data Center Dynamics / Tech Insider / Lexology
8. EU bans AI nudifiers, but enforcement remains unclear — EU lawmakers included a ban on AI “nudifiers” (tools that digitally remove clothing) in the AI Act. Enforcement mechanisms remain vague. This is one of the first concrete bans under the new AI Act and sets a precedent for harmful AI applications.
   Source: Tech Policy Press
9. DFKI and Inria establish Franco-German Center for AI — Germany’s DFKI and France’s Inria have established a joint AI research center, a milestone for European digital sovereignty. The center aims to conduct groundbreaking research and strengthen Europe’s AI position independent of US and Chinese big tech.
   Source: idw / Informationsdienst Wissenschaft
10. Netherlands blocks US acquisition of cloud provider — signal for EU digital sovereignty — The Dutch government blocked a US acquisition of a Dutch cloud provider, citing strategic digital autonomy. This is the first concrete example of a member state blocking foreign tech acquisitions on sovereignty grounds.
    Source: Jones Day / Blocks & Files

📊 Trend

From “whether we want sovereignty” to “how do we become sovereign” — urgency is palpable after the Anthropic blockade. CADA, the Tech Sovereignty Package, and national AI funds form a triptych of legislation, investment, and political momentum.