China shifts AI focus to national security after WeChat hacking warnings

China is shifting its focus on artificial intelligence risks from social harms like deepfakes to national security threats, after warnings that AI models could hack critical systems and compromise the WeChat mega-app. The country's spy agency chief said the technology could jeopardise national security. A California team demonstrated an AI-built cyberattack "worm" that could compromise millions of WeChat accounts, highlighting vulnerabilities in Chinese digital infrastructure. Beijing now sees AI safety issues as similar to those in the US, though such concerns are not publicly debated in China. Chinese policymakers are expected to discuss AI threats with US President Donald Trump at talks on Thursday, despite competition between the nations. China already regulates AI use, including banning unauthorised AI-generated likenesses, and experts say both countries face mutual exposure if AI cracks critical passwords.