Editorial illustration. Not a photograph of the leaders or the actual meeting.
By World Media Feed Editorial Team
WASHINGTON — The United States and China have agreed that humans should remain involved in military decisions involving artificial intelligence, according to Treasury Secretary Scott Bessent, offering a rare area of agreement between the two countries as they compete for technological dominance.
Bessent said the understanding was reached during President Donald Trump’s meeting with Chinese President Xi Jinping in Shenzhen.
According to Bessent, the two leaders agreed that AI systems should not be allowed to autonomously initiate major military actions.
The issue is particularly significant because Washington and Beijing are simultaneously competing over advanced AI systems, computing power and semiconductor technology.
The Trump administration continues to argue that the United States must maintain an advantage in advanced AI, while China is investing heavily in developing its own models and semiconductor capabilities.
Bessent also said Chinese AI systems had been linked to cyberattacks and argued that maintaining U.S. technological leadership was an important part of national security.
The two governments remain sharply divided over access to advanced chips.
The administration has imposed additional restrictions on China’s access to advanced AI processors, including technology used through overseas data centers.
At the same time, China is attempting to build a domestic semiconductor ecosystem capable of reducing its dependence on American technology.
The agreement on human involvement in military AI therefore exists within a much broader strategic competition.
It could nevertheless become an important principle in future discussions about military artificial intelligence.
AI systems are increasingly being developed for surveillance, intelligence analysis, logistics, cyber operations and battlefield decision support.
The central question is where governments draw the line between AI-assisted military decisions and decisions that machines can make independently.
Why it matters: As military AI becomes more capable, governments will face growing pressure to establish clear limits on autonomous weapons and other systems capable of influencing life-and-death decisions.
Source note: The understanding was described publicly by Treasury Secretary Scott Bessent. WMF is reporting his account and is not presenting it as a formal treaty unless an official agreement is released.