FENG Shuai
Deputy Director of the Institute for International Strategic and Security Studies, Shanghai Institutes for International Studies
Over the past year, global AI technology has continued to iterate at high speed. Capability breakthroughs represented by trillion-parameter agents have expanded from the digital space into physical domains such as robotics, autonomous driving, and AI-powered healthcare, as humanity enters the era of AI agents. Military threats have escalated in parallel: large language models have been integrated into operational command systems, kill chains have been dramatically compressed, and AI weapon systems including drone swarms and robotic dogs have also been combat-tested. China and the United States are engaged in systematic competition over chips, equipment, investment, and supply chains. The US employs export controls and entity lists to impose blockades, while China accelerates indigenous chip development and domestic model adaptation, with both sides forming parallel technology ecosystems. Current AI governance faces three compounding disruptions: rapid technological iteration, escalating military threats, and great power competition. The governance systems of China, the US, and the EU continue to advance in parallel without substantive convergence, while progress at the United Nations level remains slow. The overall landscape of global AI governance has not undergone fundamental change, with institutional development lagging far behind technological advancement.
