A striking dissonance now defines the global discourse on Artificial Intelligence. For years, powerful figures in Silicon Valley vehemently argued that the United States was locked in an AI arms race with China, with catastrophic consequences for the loser. However, in a significant shift this September 2026, these same tech luminaries have begun championing a different narrative: AI's potential dangers are so profound that the two global powers must urgently collaborate to decelerate the pace of development.
Anthropic's 'Pace the Frontier' Proposal Meets Chinese Skepticism
At the heart of this evolving debate is Anthropic CEO Dario Amodei, who, in his recent blog post titled 'We Must Pace the Frontier' published around September 12, 2026, argued that “Global pacing will require cooperation with China, the autocratic country with by far the most advanced AI capabilities”. Yet, Amodei's vision for a slowdown comes with a crucial caveat: a 'key part' of the strategy should be ensuring that the US and other democracies maintain “as large as possible” a lead in AI development over autocracies like China. He advocated for measures such as cracking down on chip smuggling and preventing unauthorized model distillation by Chinese companies.
'His essay specifically says the pacing of AI is conditional to increasing the lead over China,'
— Alvin Wang Graylin, Senior Fellow, Asia Society Policy Institute's Center for China Analysis
This conditional approach has rendered Amodei's proposal largely inauthentic to Chinese readers, as noted by Graylin. Beijing's response has been swift and unsparing. Guo Jiakun, a spokesperson for China’s Ministry of Foreign Affairs, described the proposal as “fear-mongering, confrontation, and vicious competition” that would only hamper global AI cooperation efforts. The Ministry further stressed China's dual emphasis on “development and security in terms of AI”.
A Clash of Priorities: Existential Threats vs. Domestic Stability
The gulf in perspectives extends beyond negotiating tactics to the fundamental nature of AI risk itself. While many prominent voices in Silicon Valley are preoccupied with so-called existential risks—the possibility of a rogue AI agent threatening humanity—Beijing's primary anxieties lie elsewhere. China's leadership is less concerned about Artificial General Intelligence (AGI) and extreme risks to humanity, and more focused on AI's potential to disrupt domestic social stability or threaten the Chinese Communist Party's (CCP) grip on power.
'The gulf lies between an Anthropic-led vision that would only subjugate and restrict China first to maximize leverage before talking. Meanwhile, China would only talk if it's being treated from a position of equals.'
— Kevin Xu, Founder, Interconnected Capital
This sentiment is echoed by China's Minister of State Security, Chen Yixin, who recently warned that AI could fuel political disinformation, cyberattacks, espionage, social unrest, and new military threats. He specifically cited US models like Anthropic's Claude and OpenAI's GPT-5.5-Cyber as posing cybersecurity risks to China. For Beijing, the primary concern is the technology's impact on political, institutional, and ideological security.
Divergent AI Risk Perceptions: US vs. China
As the AI safety community in China grows and shares many concerns with American labs, the fundamental disagreement persists: China will only engage in cooperation if it is treated as an equal partner, not as a competitor to be simultaneously constrained and subjugated for another nation's strategic advantage. This dynamic sets the stage for a prolonged, complex geopolitical dance where the future of AI development hinges on resolving deep-seated mistrust and competing national interests.
