Dario Amodei wants the global AI industry to slow work on its most powerful models. China has rejected that idea, arguing that a pause could help preserve US technological dominance rather than make the technology safer.
The debate now reaches beyond laboratories and company boardrooms. It has become part of the wider relationship between the United States and China, with both countries warning about AI risks but disagreeing over what those risks mean and who should control the rules.
Amodei’s proposal meets a Chinese wall
Amodei, the Anthropic CEO, argued that a pause would buy time to make AI safer. In a blog post, he wrote that any plan would require cooperation with China, then outlined steps aimed at slowing its AI development. Those steps included ending sales of powerful AI chips, cracking down on chip smuggling, and restricting remote access to data centers outside China.
The proposal received support from OpenAI CEO Sam Altman and xAI founder Elon Musk. Meta CEO Mark Zuckerberg and President Donald Trump rejected the idea of slowing AI development, creating a sharp split among some of the technology industry’s most prominent figures.
Trump has called fears about AI a “conspiracy” and emphasized the need for the United States to stay ahead of China. China’s officials, state media, and engineers pushed back against Amodei’s proposal, saying it treats a US advantage as if it were a neutral safety measure.
Alvin Wang Graylin, a senior fellow at the Asia Society Policy Institute’s Center for China Analysis, put the objection plainly: “I don’t see any reason why China would cooperate with [Amodei’s] proposal and be held down more than it already is.”
Two countries, two views of danger
China’s Global Times published an editorial titled “Targeting China’s AI: U.S. ‘tech right’ unfolds Cold War playbook.” On September 11, Chinese Foreign Ministry spokesperson Guo Jiakun described Amodei’s call as “fear-mongering, confrontation and vicious competition.” China also dismissed fears of existential risks as “fearmongering.”
That response does not mean Chinese officials see no danger in advanced AI. China’s national cybersecurity standards body released the third edition of its Safety Governance Framework on September 14. The document warns that AI could learn, optimize itself, and undergo recursive self-improvement at a speed or in a direction that exceeds human expectations and control.
On September 13, Chinese State Security Minister Chen Yixin wrote in China Cyberspace that US models such as Mythos and GPT-5.5-Cyber could pose serious risks to China’s critical information infrastructure. He called for tighter government oversight, greater technological self-reliance, and stronger defenses against AI-enabled attacks.
Chen also warned that AI could fabricate political rumors and incite antagonistic sentiments. In a separate warning, he wrote, “Modern warfare has entered an era increasingly shaped by artificial intelligence,” adding, “Whoever can use algorithms to achieve more refined sensing, more accurate decision-making, and more precise strikes will gain the initiative on the battlefield.”
Henry Gao, an expert on China and trade law at Singapore Management University, described the divide this way: US views often connect AI risks to catastrophic risks to humanity, while China sees the danger as geopolitical and tied to US dominance.
Pressure to move faster
Chinese engineers Tao He and Liu Shengyu publicly criticized the idea of slowing AI development. Tao He asked, “Why do you always get to define the game — and change the rules?” Liu Shengyu, an engineer at Chinese AI company DeepSeek, expressed concern about a world controlled by a few companies with powerful AI.
Huawei rotating chairman Eric Xu suggested that Chinese developers need to catch up rather than slow down. Boris But, an education entrepreneur in China, also emphasized that Chinese companies must embrace AI to avoid obsolescence.
The argument for restraint has support beyond Amodei. Otto Barten said, “AI must never get powerful enough for recursive self-improvement.” David Krueger called for more than safeguards, saying, “We need an immediate, indefinite international moratorium on frontier AI development.”
Other experts are calling for rules that can keep pace without freezing development. Wang Graylin suggested that the United States and China agree on AI capability red lines and testing standards through an independent international body. Rui Ma emphasized rules of engagement, monitoring, and safeguards that evolve alongside AI technology.
Jiang Tianjiao expressed hope that China and the United States could cooperate to manage AI risks, saying, “As such, we bear the responsibility of major powers to prevent AI from spiralling out of control, and we should set a global example by cooperating to manage these risks.”
A narrow opening for cooperation
The United States and China have not agreed on how to regulate AI or which uses should be off-limits. Mutual mistrust now shapes the argument, with Washington focused on catastrophic outcomes and Beijing worried that safety proposals could lock in a US lead.
Rebecca Bailey reported that Beijing believes slowdown calls aim to preserve US advantage and is pursuing a different balance between safety and pace. That helps explain why China can warn about recursive self-improvement, critical infrastructure, political rumors, and AI-enabled attacks while rejecting an international pause.
President Donald Trump is scheduled to meet Chinese President Xi Jinping in Washington on September 24, 2026. The two presidents are expected to agree to only limited cooperation on AI safety at the summit.
That limited opening may be the most realistic starting point. The disagreement over speed is still wide, but both countries are now describing AI as a matter of safety, security, and national power. The next question is whether those shared concerns can produce common rules before competition pushes development even faster.
Based on
- Dario Amodei wants to slow AI. China isn’t taking orders — restofworld.org
- China rejects Dario Amodei call to slow AI development ahead of Trump-Xi summit – Rest of World — restofworld.org
- As AI leaders warn of catastrophe, US and China shun slowdown calls | Technology | Al Jazeera — aljazeera.com
- Why China is pushing back on US warnings over rapid AI development | China | The Guardian — theguardian.com
