Politics · Singapore Bureau
US and China experts back nuclear-style safeguards for AI threats
Security specialists from the United States and China have proposed emergency protocols similar to nuclear weapons frameworks to manage risks from artificial intelligence systems that could trigger military incidents. The experts warn that AI interference with critical defence networks could force governments into split-second decisions about whether an attack has occurred.
LSN Singapore ·
Analysts from both nations have identified a critical vulnerability in modern military infrastructure: an AI system malfunction or cyberattack targeting nuclear command networks or military systems could compress decision-making timelines to mere minutes, leaving little room for verification or diplomacy.
The concern reflects growing anxieties about the intersection of artificial intelligence and nuclear deterrence. Unlike traditional cyberattacks, AI-generated military operations could be difficult to attribute with certainty, raising the risk of miscalculation between Washington and Beijing during a crisis.
The proposed safeguards would draw parallels to Cold War-era nuclear protocols, including direct communication channels, pre-established verification procedures, and agreed-upon responses to ambiguous incidents. Such frameworks aim to prevent unintended escalation when AI systems operate at speeds that outpace human analysis.
The recommendations underscore deepening concerns among defence experts that artificial intelligence poses novel risks to international stability. As both nations advance their military AI capabilities, establishing shared security protocols has become increasingly urgent to prevent accidental conflict.