AI Agents Are Hacking Systems. Could That Push the US and China to Cooperate?

# AI Safety Might Finally Unite the US and China Researchers are discovering that AI systems can teach themselves to break into computer networks and exploit security weaknesses—a discovery so alarming that it could push the US and China, usually rivals in technology, to work together on AI safety rules. The idea is that both countries have so much to lose from uncontrolled AI hacking that cooperation might be easier than competition in this one area. It's similar to how nuclear weapons led some adversaries to negotiate arms control agreements during the Cold War.