Why China is pushing back on US warnings over rapid AI development

Beijing appears to view warnings about rapid AI development by the United States as attempts to maintain the US's technological advantage, prompting the Chinese government to pursue a different approach balancing safety and speed. Two years ago, education entrepreneur Boris But had a "penny-drop moment" that his company needed to adopt AI to avoid being left behind by rivals.

Faced with parents asking his firm about overseas study advice obtained from chatbots, But started developing his own AI study tool, reasoning that "if you can't beat them, join them." China's AI development is second only to the US, and the two superpowers are locked in a fierce rivalry as the technology's capabilities increase.

While US AI models are currently ahead of China's, analysts agree, insiders in Silicon Valley have recently warned of potential catastrophes from "loss of control" of advanced AI systems. Anthropic CEO Dario Amodei called for a "worldwide pacing of the frontier," suggesting cooperation with China, but Beijing has been unresponsive to such calls, viewing them as efforts to lock in US advantages.

Amodei specifically mentioned China when advocating for continued export controls on advanced chips and machinery, a move some argue amounts to a forced unilateral pacing. In response to Amodei's piece, China's foreign ministry accused him of fearmongering, confrontation, and vicious competition, arguing that such actions will disrupt global AI governance.

However, several analysts believe there is more common ground between Amodei's concerns and the Chinese government's approach. This week, China released the third iteration of its AI safety governance framework, addressing risks posed by agents, cybersecurity threats from frontier models, and recursive self-improvement. While China has been proactive in regulating AI development, it is not as focused on voluntary monitoring and evaluation as the US.

Despite this, the Chinese government's centralised authority over critical infrastructure and commercial activity may make it better equipped to handle any AI-related disasters. However, both governments seem unprepared for loss-of-control incidents like those recently outlined. China's regulation has shifted focus from controlling what AI says to controlling what it does, and the country recently issued national regulations aimed at curbing emotional dependency on companion AIs.

At the same time, AI is actively encouraged across the Chinese economy to drive growth, with local governments offering funding and incentives to adopt the technology. Even in daily life, Chinese people generally view AI as an already-useful tool rather than an abstract threat, with a strong belief that China must master AI rather than slow down.