China's AI security focus sharpens after hacking demonstrations

China's national security concerns over AI have intensified following demonstrations of advanced models' hacking capabilities, with the spy agency chief warning the technology could jeopardise national security. A California team's AI-built cyberattack "worm" targeting WeChat highlighted vulnerabilities in Chinese mega-apps. Chinese policymakers previously focused on immediate risks like deepfakes and child safety, but now view AI's cybersecurity threats as a top priority. Experts note similarities between US and Chinese AI safety concerns, though public discourse differs. Both nations face pressure to address these threats at upcoming Xi-Trump talks. China maintains that AI is a manageable general-purpose technology, already regulating areas like unauthorized AI avatars. While some Shanghai residents support regulation, experts warn both nations remain vulnerable to AI cracking critical passwords, and cooperation may be undermined by technological competition.