U.S. to propose emergency AI notification system with China
The U.S. and China are racing to build the world's most powerful AI systems, and facing shared risks.
TL;DR
- U.S. Treasury Secretary Scott Bessent plans to propose an AI incident notification process with China.
- This proposal stems from China's increasing awareness of the threats posed by its own powerful open-source AI models.
- Discussions between Bessent and Chinese Vice Premier He Lifeng covered risks like uncontrolled AI agents and cyber/biological threats.
- An incident involving a Chinese AI model sending sensitive data highlighted the power and risks of these technologies.
- Bessent stressed the need for 'safe acceleration' in AI development, focusing on resilience and defense.
- He suggested that AI labs should slow down if they face issues with control over powerful models.