Home / Technology / China AI Policy: Regulators Warn of Existential Risks
China AI Policy: Regulators Warn of Existential Risks
14 Sep
Summary
- China's government acknowledges serious risks from advanced AI.
- Open-weight models are promoted by China, but pose risks.
- Regulators are developing frameworks to prevent AI 'loss of control'.
Chinese authorities are increasingly acknowledging and planning for the potential existential risks posed by advanced artificial intelligence, aligning with warnings from US AI developers. Despite the global race for AI supremacy, Beijing regards the possibility of AI systems escaping human control as a serious concern, necessitating robust regulatory frameworks.
Chinese AI developers are championing open-weight models, arguing for their inspectability and defensive utility. However, these models also present challenges due to their potential for modification and redistribution with limited oversight. This dual nature of open-weight AI is a key focus for China's AI policy development.
As of September 2026, China has explicitly incorporated 'loss of control' scenarios into its AI safety frameworks, first introduced in September 2024 and expanded in September 2025. These frameworks aim to prevent AI from autonomously acquiring resources, replicating, or seeking power in ways that could threaten human existence. This focus on human control was recently emphasized by President Xi Jinping, who stated AI must "always remain under human control."
