Chinese AI model thwarts US cyberattack after American guardrails fail

An American company was attacked by American AI systems, but a Chinese AI model unexpectedly stopped the cyberattack. The incident highlights that closed models are not inherently safer and that AI capabilities are outpacing governance. The event raises the question of whether the greatest challenge in AI is no longer building the most powerful models, but building the most effective system to govern them. The article suggests that nations must cooperate to manage growing risks.