Kevin Rudd on the Tangible Risks of US–China AI Competition

The text reports that, in an interview with The Economist, Kevin Rudd was asked about the level of the US and Chinese leaders' response to AI extinction risks. He said it was too early to judge but viewed the operationalization of the dialogue channel agreed in May (month 5) as encouraging. It says he identified agreement between the two countries on the urgency surrounding US closed frontier models and Chinese open-weight frontier models as the central issue. He points out that China is optimistic about cyberdefense because of its confidence in its firewall, but that this is a mistake because it would be insufficient against an all-out attack on infrastructure if AI escaped control. The text reports that neither side has reached an internal consensus because of disagreements between the technology community and policy and political circles, and that China's technology community is concerned about the proliferation and misuse of autonomous agents amid advances between May (month 5) and September (month 9). It introduces a question about whether Trump's insistence on US superiority and Dario Amodei's position on containing China are obstacles preventing China, as the country playing catch-up, from participating in negotiations. It says China's basic position is also to compete for victory across the entire AI stack, and that, after reviewing guardrail proposals and Amodei's writing, China is holding firm that it will not tie its own hands. It concludes by presenting the dilemma of whether cooperation will require a disaster on the scale of the Cuban Missile Crisis or whether policymakers will heed technologists' warnings proactively, and expresses optimism about making the negotiating track operational on the basis of China's science-oriented culture. All outlooks expressed are the speaker's conditional judgments, not definitive short-term price predictions.