US-China Experts Push Nuclear-Style AI Safeguards — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Melanie Sisson (Brookings Institution) and Tianjiao Jiang (Fudan University) published joint proposals urging U.S.-China red lines on AI interacting with nuclear command systems, released last week ahead of a planned September 24 Trump-Xi meeting in Washington and government-level AI talks.
- Sisson proposed humans retain sole authority to initiate AI-enabled cyberattacks against the other country's nuclear command, control and communications systems or strategically important infrastructure, warning autonomous systems could malfunction or be compromised.
- Jiang proposed explicit red lines preventing AI from independently deciding to use nuclear weapons or autonomously attacking nuclear command systems, and urged a shared definition of "meaningful human control" to prevent both countries from using identical language to justify different machine-autonomy levels.
- Jiang also proposed a dedicated U.S.-China military hotline for AI incidents, arguing automated cyber exchanges could escalate too quickly for human intervention — one country's defensive AI might respond to suspicious activity while the other interprets it as an attack.
- Carla Freeman (Johns Hopkins SAIS) questioned the utility of the existing U.S.-China military crisis hotline, noting Chinese officials refused to answer U.S. counterparts' calls about the February 2023 spy balloon incident, and that China's top-down bureaucracy disempowers low-ranking officers from initiating communication.
- China has placed AI inside its Foreign Ministry's Department of Arms Control, while Washington has no equivalent single arms-control home — policy is spread across the White House National Security Council, State Department, Pentagon and other agencies.
- Both governments remain wary of self-restraint: Trump has warned broad AI restraints could hand China an advantage, while Beijing views U.S. technology controls as attempts to preserve American dominance.
Why it matters: The proposals land just days before the September 24 Trump-Xi summit, but Freeman's documented skepticism about the existing military hotline — based on China's refusal to engage during the 2023 spy balloon episode — suggests any new AI incident protocol will inherit the same bureaucratic and political obstacles that have already stalled crisis communication between the two nuclear-armed rivals.
Ask SkimNews
