Anthropic CEO Dario Amodei’s call to slow the development of artificial intelligence has met immediate pushback from China, which sees the proposal as a bid to preserve U.S. technological dominance rather than a genuine safety measure.
The Proposal and Global Reactions
Amodei outlined his plan earlier this month, urging the global AI industry to pause development of its most powerful models to focus on safety. He argued that such a pause would buy time to address risks and require cooperation with China, which he described as having “by far the most advanced AI capabilities.” To slow China’s progress, Amodei proposed restricting the sale of AI chips and semiconductor equipment, cracking down on smuggling, and limiting remote access to data centers outside China.
The idea gained unexpected traction among U.S. AI leaders. OpenAI CEO Sam Altman and xAI founder Elon Musk voiced support, while Meta CEO Mark Zuckerberg and former President Donald Trump dismissed it, advocating for unrestricted AI growth. However, China’s response has been unequivocally negative.
China’s Rejection and Counterarguments
Chinese officials and state media swiftly rejected Amodei’s proposal, framing it as a strategic move to contain China’s AI advancement. The Global Times published an editorial titled “Targeting China’s AI: U.S. ‘tech right’ unfolds Cold War playbook,” arguing that the initiative aimed to limit Chinese innovation to maintain American supremacy. On September 11, Foreign Ministry spokesperson Guo Jiakun labeled Amodei’s call “fear-mongering, confrontation and vicious competition,” warning it would undermine global AI governance.
Alvin Wang Graylin, a senior fellow at the Asia Society Policy Institute, noted that China has no incentive to cooperate with Amodei’s terms, as the restrictions would further limit its access to essential technologies. Graylin acknowledged that China might eventually agree to pace AI development but emphasized it would not accept Amodei’s specific framework.
China is simultaneously advancing its own AI safety standards. Its national cybersecurity body released a third edition of its Safety Governance Framework on September 14, warning that AI could autonomously improve itself beyond human control. This mirrors concerns raised by former Anthropic researcher Jacob Coxon about existential risks.
Zilan Qian, a research scholar at GovAI, observed that while China and the U.S. share similar worries about AI risks, their political and socioeconomic contexts shape different priorities. Chinese policymakers are more focused on issues like AI’s impact on labor, misinformation, and cybercrime, whereas U.S. discussions often center on existential threats.
Engineers and Industry Voices
Chinese AI engineers have largely remained silent on Amodei’s proposal, but some have publicly criticized it. Tao He, head of R&D at MiniMax Code, questioned why China should accept restrictions when the U.S. is not slowing down. “Oh, come on. Yesterday, the mission was AGI. Today, we’re told to slow down,” He wrote on social media, challenging the notion of who sets the rules.
Liu Shengyu, a DeepSeek engineer, described a dystopian vision where a few companies control powerful AI, exacerbating inequality. He advocated for making advanced AI widely accessible, expressing distrust in Anthropic and OpenAI to manage such power responsibly.
Chinese State Security Minister Chen Yixin, in a September 13 article, highlighted AI’s role in modern warfare, asserting that mastering algorithms for precision strikes grants battlefield superiority. He called for stricter oversight and technological self-reliance.
Huawei’s rotating chairman, Eric Xu, suggested that U.S. labs may face risks Chinese developers cannot yet perceive.
This diplomatic engagement comes amid escalating tensions over Amodei’s proposal and China’s simultaneous advancement of its own AI safety initiatives.
China’s Accelerated Safety Framework
China’s national cybersecurity standards body released a third edition of its Safety Governance Framework on September 14, explicitly warning that AI could autonomously learn, optimize itself, and undergo “recursive self-improvement” at speeds exceeding human expectations and control. The framework represents an expansion of Beijing’s domestic approach to AI governance, addressing concerns that mirror the dystopian scenarios previously highlighted by former Anthropic researcher Jacob Coxon.
Zilan Qian, a London-based research scholar at GovAI, observed that while China and the U.S. share similar fundamental concerns about AI risks, their distinct political and socioeconomic contexts create different priorities.
Defense and Self-Reliance Concerns
Chinese State Security Minister Chen Yixin warned that advanced U.S. models such as Anthropic’s Mythos and OpenAI’s GPT-5.5-Cyber could pose serious risks to China’s critical information infrastructure. In a September 13 article published in the state-run magazine China Cyberspace, Chen emphasized that modern warfare has entered an era increasingly shaped by artificial intelligence, asserting that whoever can use algorithms to achieve more refined sensing, more accurate decision-making, and more precise strikes will gain the initiative on the battlefield.
His warnings reflect Beijing’s view that AI mastery is essential for military and national security advantages in future conflicts.