China Steps Up AI Safety Measures Amidst Growing Loss-of-Control Concerns

President Xi Jinping has emphasised that AI should remain under human control

China is strengthening measures to address the potential risks posed by increasingly powerful artificial intelligence systems, following growing warnings from researchers in the United States, that advanced AI could eventually become difficult to control. As the United States and China lead global development of frontier AI, Chinese policymakers have increasingly focused on the possibility of advanced systems operating beyond effective human oversight.

China’s AI safety framework warns that future AI could potentially acquire resources, replicate itself and seek greater control, creating risks to human safety and development. Chinese officials have also raised concerns about the security implications of advanced AI models, particularly their potential impact on critical infrastructure and military systems.

President Xi Jinping has emphasised that AI should remain under human control, while Chinese authorities have called for caution in the development and deployment of military AI. Beijing has begun translating these concerns into regulations for increasingly autonomous AI agents.

Guidelines issued require developers to strengthen systems for detecting, intervening in and blocking harmful autonomous behaviour, while ensuring that humans retain final decision-making authority.