Three of America’s most powerful tech CEOs have, within a week, publicly argued that the United States should not block Chinese AI models – a striking break from Washington’s push to contain them.
Meta’s Mark Zuckerberg told the Financial Times on July 29 that the U.S. should not bar Chinese AI models from the race for advantage, warning that America’s frontier labs could use “regulatory capture” to smother domestic competition. Banning cutting-edge Chinese AI at home, he argued, “is not an effective solution.”
Tesla’s Elon Musk said on July 27 that the probability of a Chinese AI company becoming the leader at some point is “quite high,” and opposed proposals to forbid U.S. firms from using Chinese models because “it won’t work, and it won’t stop China from becoming an AI leader.” His point: AI’s real constraints are electricity and AI chips, and China holds clear advantages in both.
Nvidia’s Jensen Huang told Axios on July 22 the U.S. should not ban or restrict Chinese models like Kimi K3. “These Chinese models are excellent,” he said, and open models should be used. He argued the market misread DeepSeek’s impact the first time, and is misreading Kimi’s now. More open models, wherever they come from, expand the AI market and pull through more chips, data centers and compute demand.
The rare alignment of three CEOs who normally compete fiercely is itself the story. As Washington weighs export controls and model bans, the people building the most advanced AI infrastructure in the West are betting that Chinese open-source models are a net positive – and that trying to wall them off would only hurt U.S. competitiveness.
Source (Chinese original): Sohu IT
Translated and adapted from Sohu IT (it.sohu.com).