Key facts
- China views AI development as a core sovereignty capability and is not receptive to US-led slowdowns.
- China's AI development pace and quality are considered second only to the United States.
- US models are still considered ahead of Chinese counterparts by analysts.
- China released its third AI safety governance framework, addressing risks from AI agents and self-improvement.
- US President Donald Trump dismissed AI existential risk fears as a 'hoax'.
China is resisting calls from the US and Silicon Valley insiders for a slowdown in advanced artificial intelligence development, viewing such proposals as an attempt to preserve American technological superiority. While analysts generally agree that cutting-edge US AI models remain ahead of their Chinese counterparts, China is accelerating its progress, framing AI as a critical sovereign capability.
Anthropic CEO Dario Amodei recently suggested a global pacing of frontier AI development, explicitly mentioning the need for cooperation with China. However, technology research group Omdia noted that China is "definitely not receptive" to a slowdown dictated by US terms, as Beijing aims to close the gap with US models. Amodei also advocated for continued export controls on advanced chips to China.
In response to Amodei's piece, China's foreign ministry stated that 'fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance.' However, some analysts suggest this does not indicate a rejection of broader AI safety concerns. China has released its third AI safety governance framework, which maps AI risks and potential mitigation strategies, including those related to AI agents and recursive self-improvement. Concordia AI researcher Gabriel Wagner believes there is more common ground on AI safety between China and the US than commonly assumed.
This contrasts with US President Donald Trump, who dismissed fears of AI posing an existential threat as a 'hoax.' Trump is expected to discuss AI with Chinese President Xi Jinping during an upcoming meeting.
China's government, concerned about political control, has been proactive in regulating AI development. Omdia's Lian Jye Su suggested that China's centralized authority might position it better than the US to handle potential AI disasters. However, the Carnegie Endowment for International Peace's Leia Wang noted that Chinese companies conduct less voluntary monitoring than their US counterparts, and neither government appears fully prepared for a loss-of-control incident.
China's regulatory approach is described as iterative, with rapid adaptation and new measures. The country has had an algorithm registry since 2022, and recent regulations focus on controlling AI actions rather than just its output. Beijing is also actively encouraging AI adoption across its economy to drive growth, with local governments offering incentives to companies. For instance, Black Lake Technologies reported working with 40,000 factories globally on industrial AI agents, where reliability is a key concern for manufacturers. Daily life in China is highly digitized, with a practical attitude towards AI as a useful tool rather than an abstract threat, despite anxieties about jobs and misinformation.