Ahead of Sam Altman's UN address, OpenAI urges the US to lead an effort to develop global safety and security standards for building frontier systems

Governments now propose a shared system to report national security-level AI incidents.

U.S. Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng discussed setting up an AI dialogue mechanism and a new "notification mechanism" for AI incidents that could impact national security. This discussion occurred ahead of a meeting between U.S. President Donald Trump and Chinese leader Xi Jinping, amidst a broader U.S.-China competition for AI dominance and rising global concerns about AI threats. The U.S. proposed the notification system to share information on AI incidents, particularly those affecting national security. Discussions also touched upon shared risks like AI-enabled cyberattacks and biological misuse, with both nations expressing concerns about AI in the other's hands. Despite geopolitical tensions, there is recognition of potential cooperation on specific AI risks, although U.S. export controls on advanced AI technologies remain a point of contention. China is increasingly developing competitive frontier models despite these restrictions, and accusations of "distillation" of U.S. AI technology by Chinese labs are a sore spot.

The U.S. proposal for an AI incident notification mechanism signals a pragmatic approach to de-escalating potential AI-related international crises. This move indicates a recognition that complete AI decoupling is not feasible and that some level of shared situational awareness is necessary, even between geopolitical rivals. The focus on "national security level" incidents suggests that dialogue will likely center on high-impact events such as AI-powered cyberattacks on critical infrastructure or the potential for AI in weapons systems.

The ongoing competition for AI dominance, however, complicates these safety discussions, with U.S. export controls on advanced chips and accusations of technology "distillation" creating friction. China's progress in developing competitive frontier models despite these restrictions highlights the challenges of containing technological advancement. Future cooperation will likely be shaped by a delicate balance between national security concerns, economic competition, and the shared existential risks posed by advanced AI.