Anthropic CEO urges AI ‘disarmament’ talks in Beijing

A Silicon Valley CEO walked into Beijing and dared to reframe the AI arms race as an arms-control problem. On September 13, 2026, Anthropic’s Dario Amodei urged disarmament-style talks between the U.S. and China to slow frontier AI, positing verification regimes and capability thresholds as the only brakes strong enough to matter. In a week where export controls, access fights, and model races dominate headlines, he pitched strategic stability over sprint-speed supremacy.

The sky matches the script: a Waxing Crescent in Libra pushing consensus-building while the Sun locks a precision sextile to Mars—less grandstanding, more tools-on-the-table problem solving. But the Moon’s square to Mars signals hot reactions, and Mercury’s tangle with Neptune warns of fog in the message. The promise: technical guardrails drafted at pace. The peril: public blowback and definitional warfare. Our forward-looking thesis: Narrow, testable confidence-building measures are favored first—if rhetoric can survive the next 72 hours of heat and the next two weeks of definitional fights.