On Wednesday, a researcher named Jacob Coxon quit his job at Anthropic and posted a warning that stopped the internet cold. "Those who are building AI genuinely believe it could kill all of us before the decade is out," he wrote. By Monday, his words had already crossed the Pacific and ignited a diplomatic row. Beijing accused Washington of weaponising doomsday rhetoric — and suddenly, the global debate about AI safety had become the latest front in the fiercest geopolitical rivalry of our era.

Donald Trump moved fast to define the battle lines. "Whoever wins AI will be the winner," he said on Sunday. A day later he sharpened his aim: the only party pleased by any slowdown in AI development, he argued, was China. Beijing responded with fury, charging the United States with recycling Cold War language.

Rarely have the two superpowers made it so nakedly clear that regulating artificial intelligence is now inseparable from dominating the world order. The confrontation echoes a fear that haunted an earlier American generation — the sight of a Soviet satellite, Sputnik, blinking overhead in 1957 and proving that a rival could reach places the United States could not. That shock sent America to the Moon. This time, the stakes stretch across chip supply chains, armies of software engineers, and the rules that will govern a technology already reshaping economies on every continent, from São Paulo to Seoul.