US export controls force Anthropic to shut down advanced AI models worldwide, triggering debate in Europe over whether to build homegrown alternatives or secure access through trade policy.

What happened: The US government issued an export control order that forced Anthropic to pull its Fable 5 and Mythos 5 models from worldwide access, blocking non-US citizens from using them. The European Commission is assessing the practical impact, and talks to restore access are ongoing. Anthropic CEO Dario Amodei will join other AI company leaders at a G7 working dinner on Wednesday.

Why it matters: A single foreign government order shut down critical technology instantly for European companies and researchers, raising questions about dependency on US-controlled AI systems during geopolitical tensions. European experts agree this is a wake-up call for technological sovereignty, but they sharply disagree on the solution—some argue Europe must build its own capable models and infrastructure, while others contend that approach is unrealistic and that Europe should instead negotiate contractual access backed by trade policy threats.

What to watch: The structural barriers are substantial. Researchers note Europe would need not just one but several homegrown AI providers, large-scale data centers, and sufficient power generation—with one noting that German power capacity has dropped back to 1985 levels. The outcome of talks between Anthropic and the US government on restoring model access will signal whether commercial negotiation or regulatory intervention shapes the path forward.