China Isn’t Buying Silicon Valley’s Call for an AI Slowdown
Silicon Valley leaders are urging cooperation with China to slow AI development due to potential dangers, but China views the proposal skeptically, perceiving it as an attempt to maintain US dominance.
Intelligence analysis by Gemini 2.5 Flash
US tech executives, including Anthropic CEO Dario Amodei, are advocating for a global AI slowdown, emphasizing the need for cooperation with China. However, Beijing and Chinese AI companies are wary, interpreting the call as a strategy to restrict China's progress and secure a US lead, rather than a genuine safety initiative.
Imagine two kids are building the fastest race cars ever. One kid, from America, says, "These cars are getting super fast, maybe too fast! Let's slow down together before someone gets hurt." But the other kid, from China, thinks, "Wait, you just want me to slow down so *you* can stay ahead and win the race later!" They both agree fast cars can be risky, but they don't trust each other's reasons for wanting to slow down.
Analysis
Dario Amodei's Proposal
Anthropic CEO Dario Amodei recently published a blog post advocating for a global AI slowdown, emphasizing the urgent need for cooperation with China. His argument centers on the potential dangers of advanced AI, suggesting that a collaborative effort is essential to manage the technology's rapid progression. However, a critical component of Amodei's proposed strategy involves the US and other democracies maintaining "as large as possible" a lead in AI development over autocratic nations like China.
This condition has been met with significant skepticism from Chinese stakeholders. Amodei's call for a slowdown, coupled with recommendations to crack down on chip smuggling and prevent unauthorized model distillation by Chinese companies, is perceived in China as an attempt to restrict their progress rather than a genuine plea for equitable safety collaboration. The inherent contradiction of seeking cooperation while simultaneously aiming to widen a technological gap undermines the authenticity of the proposal in the eyes of Chinese readers and policymakers.
Guo Jiakun's Response
China's Ministry of Foreign Affairs spokesperson, Guo Jiakun, publicly dismissed Amodei's proposal, characterizing it as "fear-mongering, confrontation, and vicious competition." This strong reaction underscores Beijing's resentment towards being simultaneously asked to be an essential governance partner and treated as a competitor requiring constraint. The Chinese government has consistently stressed its commitment to balancing "development and security in terms of AI," indicating a different set of priorities than those articulated by some Silicon Valley leaders.
Beijing's official stance reflects a broader skepticism about the motivations behind such calls for a slowdown. Chinese policymakers tend to view these proposals instrumentally, suspecting that US companies might be seeking to inflate their own value or slow down their primary competitors, which are increasingly Chinese firms. This deep-seated distrust makes it challenging for any US-led initiative to gain traction without a fundamental shift in approach that acknowledges China as an equal partner.
Michelle Nie's Analysis
Michelle Nie, a visiting fellow at the Center for a New American Security, offers crucial insight into Beijing's perspective, noting that Chinese leaders are not primarily concerned with the "existential risks" that preoccupy many US frontier AI labs. Instead, Beijing's anxieties are predominantly focused on how AI could disrupt domestic social stability or threaten the Chinese Communist Party's grip on power. This includes concerns about political disinformation, cyberattacks, espionage, social unrest, and new military threats, as highlighted by China's minister of state security, Chen Yixin.
Nie's analysis suggests that Beijing finds it difficult to take US calls for AI safety at face value partly due to skepticism about achieving Artificial General Intelligence (AGI) and doubts about narratives of extreme risks to humanity. While China is beginning to recognize "loss-of-control risks," the paramount concern remains the technology's potential impact on CCP rule. This fundamental divergence in perceived risks and priorities creates a significant gulf in understanding and makes genuine, trust-based cooperation on AI governance a complex challenge.
Key points
- Silicon Valley leaders are advocating for a global AI slowdown and cooperation with China due to perceived dangers.
- China views these proposals skeptically, interpreting them as attempts to restrict its AI development and maintain US leverage.
- Anthropic CEO Dario Amodei's proposal explicitly suggests increasing the US lead over China as a "key part" of the slowdown strategy.
- Chinese officials and researchers are concerned about AI's impact on domestic social stability and CCP rule, rather than existential risks.
- Many in China's AI industry prefer open, inexpensive models to prevent a few companies from controlling advanced AI.
Despite current skepticism, the article notes that preliminary negotiations between the US and China on AI safety are underway, suggesting a potential path for future cooperation if trust can be built and proposals are framed as equitable partnerships.
The deep distrust and differing priorities, with the US seeking to maintain a lead and China prioritizing social stability over existential risks, could lead to a continued AI arms race without meaningful global safety agreements.
