Around 100 countries across the globe attended the Responsible AI in the Military Domain (REAIM) summit in Seoul where the agreement was discussed. China was the only country that didn't sign the (non-binding) treaty to prevent that artificial intelligence makes the key decisions on using nuclear weapons.
#China refuses to sign agreement banning AI from controlling nuclear weapons at REAIM summit
1 messages · Page 1 of 1 (latest)
I feel like this is a catch-22 iteration of historical M.A.D. Policy. Now it’s not just nuclear weapons that are necessary for deterrence, but AI-controlled nuclear weapon systems as well.
If China wants to start off that specific arms race, there are only two possible outcomes I see.
1). Global acceptance of preemptive authorization and strikes against weapon systems “gone rogue” through automation rather than state control. Unless direct channels of communication are opened between these systems and foreign countries, this could be seen as the equivalent of a nation giving nuclear weapon capabilities to unauthorized third parties and must be treated as such.
2). Global acceptance that China is setting the new standard for automated nuclear deterrences and the rest of the world needs to follow in their footsteps. This scenario worries me more, as the first scenario can still occur, but with even more players involved if this isn’t shut down immediately.