G7 Évian — AI sovereignty fracture surfaces
The G7 summit concluded without a joint communiqué (none was planned given US–European tensions). Key outputs: Ukraine support reaffirmed; nine declarations adopted on AI, cancer, Ebola, and critical minerals. Macron hosted AI CEOs including Dario Amodei (Anthropic), Sam Altman (OpenAI), and Demis Hassabis. A “trusted partners” framework for selective access to advanced US AI models was discussed — driven directly by the June 13 export control order blocking all foreign national access to Anthropic’s Fable 5 and Mythos 5.

New today: Euronews confirmed the export control issue “loomed large” in AI working sessions — G7 allies realised they could be cut off from frontier AI “at a moment’s notice.” Trump threatened France with 100% tariffs on wine over digital services taxes.
Why it matters: ⚑ The Anthropic export control event has become a civilisational signal. It demonstrated in real time that frontier AI can be treated as a weapons-adjacent export — similar to nuclear or semiconductor controls. European and Canadian leaders (Carney explicitly) are treating it as a structural lesson in AI sovereignty risk. The “trusted partners” framework, if it emerges, would create a two-tier global AI access regime with long-term consequences for non-allied AI development and banking-sector technology dependence.
Sources: WION/AI News · Euronews

https://www.euronews.com/my-europe/2026/06/17/six-takeaways-from-the-g7-summit-in-evian