Trump's AI Safety Answer: Slow China, Not America — SkimNews
Get the Geopolitics newsletter
Daily geopolitics — wars, elections, sanctions, the diplomatic moves that move markets. Free.
- Anthropic CEO Dario Amodei published "We Must Pace the Frontier" on September 12 urging leading AI labs to deliberately slow capability gains, prompting Sam Altman to commit OpenAI to host independent evaluators and Elon Musk to reply "Dario is right."
- President Trump called fears of an AI catastrophe a hoax two days after Amodei's essay, while Treasury Secretary Scott Bessent warned there is "no day after tomorrow if China wins" on AI and "nothing [else] would matter."
- The Trump administration is embracing the "slow China" half of the AI industry's agenda while rejecting the "slow ourselves" half, focusing on chip export controls and distillation enforcement rather than constraining U.S. labs.
- American public opinion leans toward slowing AI: a Politico poll found 63% see a real chance AI could wipe out humanity, and 2024 Trump voters split 50% to 38% in favor of staying in the AI race over prioritizing safety.
- NSA, FBI, and CISA jointly accused six Chinese developers, including DeepSeek and Alibaba, of industrial-scale distillation of American frontier models earlier this month, and the Trump administration has cleared roughly 10 Chinese firms to buy Nvidia H200 chips that customs then blocked.
- Bessent proposed a standing China-U.S. AI dialogue in New York on September 20 with a mechanism for flagging AI incidents that threaten national security—the first such proposal of Trump's second term, though Beijing's response was unclear.
- Chinese AI governance figures Xue Lan and Zeng Yi, who helped draft China's national AI governance principles, joined a Capitol Hill discussion convened by Senator Bernie Sanders this spring and called for "safe zones" where American and Chinese scientists build joint safety standards.
Why it matters: The Trump administration's strategy of restricting Chinese AI while letting U.S. labs operate unconstrained satisfies midterm voters anxious about AI risk, but risks poisoning a potential China-U.S. incident notification channel before it starts—Beijing will read "AI safety" as containment, eroding trust exactly when the channel matters.
Ask SkimNews
