World leaders want American AI, just not America's kill switch

World leaders want American AI. They just don’t want America to be able to turn it off.

At the G7 summit, Macron and Modi flagged the risk that the U.S. could cut off access to American AI models overnight. The fear got real after a recent Anthropic outage locked European users out of Claude. Macron told Amodei, Altman, and Trump over lunch that no country can wire critical infrastructure into a model the U.S. can switch off at will. The post doesn't lay out specific policy proposals, but it nails the tension: American AI is the best, but depending on it is a sovereignty gamble.

Why it matters: Macron and Modi raised the kill-switch risk directly with US AI execs and Trump at G7, anchored by a concrete incident (Anthropic outage). The geopolitics + AI infrastructure dependency angle hits practitioners directly. Downside: the piece stops at describing the phenomenon, ...