China Isn’t Buying Silicon Valley’s Call for an AI Slowdown

Calls for global AI slowdowns are being met with accusations of strategic maneuvering, not genuine safety concerns, from China.

Anthropic CEO Dario Amodei proposed a global AI slowdown requiring cooperation with China, emphasizing that the US and its allies should maintain a significant lead over autocratic nations. He suggested measures like cracking down on chip smuggling and preventing unauthorized model distillation by Chinese companies to achieve this. China's government and AI industry have responded with skepticism, viewing Amodei's proposal as an unfavorable "bad deal." Chinese officials characterize the suggestion as "fear-mongering" and "confrontation," asserting that China prioritizes both AI development and security equally. Chinese entities are wary of US intentions, suspecting that such calls for safety and slowdowns are strategic moves to inflate US companies' value or hinder their competitors. While China shares some concerns about AI risks, its primary anxieties focus on domestic social stability and the ruling party's grip on power, rather than existential threats.

China perceives proposals for an AI slowdown and safety cooperation as a tactic by US companies to gain a competitive advantage and inflate their market value, rather than a genuine effort to mitigate existential risks. The core of China's skepticism lies in the belief that the US seeks to restrict its development while simultaneously demanding collaboration, a proposition viewed as inequitable. This dynamic suggests that international AI governance discussions will be heavily influenced by geopolitical competition and mutual distrust.

The differing priorities between US and Chinese AI safety concerns—existential threats versus domestic stability and political control—create a fundamental disconnect. While US labs focus on long-term, civilization-level risks, Beijing is more preoccupied with AI's immediate impact on social order and the Communist Party's authority. This divergence means that any future cooperation will likely require navigating these distinct anxieties and finding common ground on narrowly scoped issues that address both perspectives.