Emergency Response Measures for Catastrophic Risk

James Zhang, Miles Kodama, Zongze Wu, Michael Chen, Yue Zhu, Geng Hong

Keywords: China AI Governance, Emergency Response, Catastrophic AI Risks, Frontier Safety Policy

TL;DR: Conceptualizing and recommending emergency management measures for catastrophic AI risks in China.

Abstract:

Chinese authorities are in the process of extending the county’s four-phase emergency response framework (prevent, warn, respond, recover) to manage risks from advanced artificial intelligence. A range of concrete implementations for the proactive prevention and warning phases are still being explored. This paper analyzes an implementation model drawn from worldwide AI safety best practices: frontier safety policies (FSPs). The key features of these policies are pre-deployment evaluation for dangerous capabilities and tiered, pre-planned security and safety
measures. We observe close alignment between FSPs and the proactive phases of China’s emergency response framework. This suggests that China could use the FSP model to operationalize emergency response in a manner consistent with China’s own established governance principles.

Submission Number: 23