As G7 leaders gather in France for the third day of their summit to discuss artificial intelligence, fears of a potential American “kill switch” have rarely felt more real. Access to AI can be restricted overnight, as was made clear when the Trump administration cut off foreign access to Anthropic's most powerful models. So can Europe afford to rely on technology it does not control? And does it need its own AI, or is that unrealistic? At the G7, leaders meet the CEOs shaping the future of AI — but the bigger question remains unanswered.