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

China is dismissing Silicon Valley's plea for an artificial intelligence (AI) slowdown, according to a report. Anthropic CEO Dario Amodei wrote in a blog post that global pacing requires cooperation with China, which possesses the most advanced AI capabilities, but Beijing and its AI companies are skeptical of this proposal. The issue lies in their view that they are offered a bad deal, as they believe they are being offered a subjugated and restricted position first, before any talks.

Kevin Xu, founder of Interconnected Capital, explains that China would only engage if it's treated as an equal partner. Amodei urged the US to maintain a substantial lead in AI development over autocracies like China by curtailing chip smuggling and preventing unauthorized model distillation by Chinese firms. However, many Chinese officials, AI researchers, and commentators find this proposal inauthentic and resent the treatment of their country as both an essential governance partner and a competitor that needs to be constrained.

China's Ministry of Foreign Affairs dismissed the proposal as "fear-mongering, confrontation, and vicious competition" that would hamper global AI cooperation. Chinese leaders are primarily concerned about AI disrupting social stability or threatening the Communist Party's grip on power, rather than existential risks such as the possibility of a rogue AI agent killing everyone on Earth.

The government's skepticism is also shared by many people within China's AI industry, which focuses on releasing lower-cost, open-weight models. While there are signs of preliminary negotiations between the US and China, both governments share concerns about cybersecurity threats, misuse of powerful models, and potential unexpected behavior from increasingly capable AI systems.