Anthropic CEO Dario Amodei is calling for US-China cooperation on AI safety, though his proposal to maintain American technological dominance has faced significant skepticism from Chinese officials.

Key Points

• Anthropic CEO Dario Amodei published a blog post urging the US and China to collaborate on slowing the development of dangerous frontier AI models.
• Amodei’s strategy includes maintaining a US lead by restricting Chinese access to advanced chips and preventing unauthorized model distillation.
• China’s Ministry of Foreign Affairs rejected the proposal, labeling it as "fearmongering" and a form of "vicious competition" rather than genuine cooperation.
• Chinese leadership prioritizes AI risks related to social stability and Communist Party control over the existential "doomsday" scenarios often cited by Silicon Valley.
• Researchers at Chinese firms like DeepSeek argue that US tech companies are using safety narratives to stifle competition and consolidate market power.
• Preliminary diplomatic negotiations regarding AI safety are reportedly underway ahead of a planned meeting between President Donald Trump and Xi Jinping.

Why it Matters

The tension between US tech leaders and Beijing highlights the difficulty of establishing global AI governance when national security and economic competition remain primary drivers. If these diplomatic efforts fail, the lack of a unified safety framework could increase the risk of uncontrolled AI development and geopolitical instability.