China refuses to sign agreement to ban AI from controlling nuclear weapons

China opted out of the 'Blueprint for Action' agreement which seeks to ban artificial intelligence from controlling nuclear weapons. The agreement was adopted at the Responsible AI in the Military Domain (

REAIM

) summit in

Seoul

on Tuesday where over 100 countries including the US were present.
The agreement is not legally binding and it seeks to "maintain human control and involvement for all actions...

concerning nuclear weapons employment."
"AI applications should be ethical and human-centric," it said.
Calling the AI a "double-edged sword," South Korean defence minister

Kim Yong-Hyun

said, "As AI is applied to the military domain, the military's operational capabilities are dramatically improved. However it is like a double-edged sword, as it can cause damage from abuse."

The declaration from the summit did not specify sanctions or penalties for violations.
It acknowledged that significant progress is needed for states to keep up with advancements in military AI, emphasizing the need for further discussions to establish clear policies and procedures.
The Seoul summit, co-hosted by Britain, the Netherlands, Singapore, and Kenya, builds on the inaugural event held in The Hague in February last year. It positions itself as the "most comprehensive and inclusive platform for addressing AI in the military domain."
Russia was excluded from the summit due to its ongoing invasion of

Ukraine

.