How China Is Preparing Against the Risk of Artificial Intelligence Escaping Human Control
Beijing is increasing national security measures by taking into account warnings from US developers that advanced models could escape human control.
Warnings from US artificial intelligence developers that advanced models could escape human control and lead to the destruction of humanity have drawn attention in China. High-level officials and policymakers are preparing comprehensive security strategies against these existential risks.
AI Development by the US and China
The United States and China stand out as two major powers in the development of frontier artificial intelligence technologies.
Warnings from China's Minister of State Security
Chinese Minister of State Security Chen Yixin stated that advanced US models such as Anthropic's Mythos and OpenAI's GPT-5.5-Cyber could pose serious risks to the country's critical information infrastructure.
Minister Chen Yixin called for a comprehensive reinforcement of artificial intelligence security against these threats.
Security Framework and Legal Steps
China first included the scenario of artificial intelligence going out of control in a security framework published under the guidance of the Cyberspace Administration of China in September 2024.
In September 2025, the Beijing administration shared an expanded version focusing on preventing loss of control with the public.
Artificial Intelligence Conference in Shanghai
President Xi Jinping emphasized at the World Artificial Intelligence Conference held in Shanghai that artificial intelligence must always remain under human control.