Mark Zuckerberg says the US should not block Chinese AI models
Meta CEO Mark Zuckerberg has publicly argued against the idea that the United States should keep advanced Chinese artificial intelligence out of its market, calling the approach “not an effective solution” in an interview with the Financial Times published this week. His reasoning is competitive rather than sentimental: instead of restricting Chinese models, Zuckerberg said, the US should identify where it is actually behind and fix those bottlenecks. The position puts him alongside startups that have urged the Trump administration not to ban Chinese open-weight models, and it lands in the middle of a live policy fight in Washington over export controls and model restrictions.
What Zuckerberg said about Chinese AI
Speaking to the Financial Times, Zuckerberg argued that restricting Chinese models would not deliver the strategic outcome Washington wants. “The big trend in the industry’s development has always been towards greater openness, putting technological power in the hands of more people, not fewer,” he told the paper. He framed openness as the natural direction of travel for AI, and suggested that a closed, gatekept alternative concentrates power in too few hands. His remarks implicitly criticised OpenAI and Anthropic, which have lobbied for tighter restrictions on the most capable so-called frontier models.
How the industry has split on open versus closed AI
US AI companies have divided into two camps on the question. Meta, Microsoft, Nvidia, and Elon Musk have backed open-weight models, the term for systems whose trained parameters are publicly released so others can run and modify them. OpenAI and Anthropic have argued for stricter controls over the most capable systems, citing safety and national-security concerns, including the risk that powerful models could enable cyber-attacks or bioweapon development. Zuckerberg’s side sees a different primary danger: that a small number of American labs, working with sympathetic regulators, end up controlling a technology they argue should be widely distributed, and that the resulting slowdown could be justified in the name of protection.
The Chinese models sharpening the debate
Chinese AI progress has given the argument urgency. The article notes that a cheap Chinese model has been closing in on American leaders, and that the release of Moonshot AI’s Kimi K3 in mid-July sharpened the question of what, if anything, US policymakers should do in response. On 24 July, twenty-five organisations signed an open letter titled “Open Weights and American AI Leadership,” arguing that openness, not restriction, is the surer route to continued US primacy in AI. Nvidia CEO Jensen Huang has also pushed back against hawks in Washington who favour tighter restrictions.
Why Meta’s stance is awkward
There is an obvious tension in a Meta executive making the case for openness. The company built its AI reputation on the open-weight Llama family, yet its newest flagship, Muse Spark, arrived closed source, a shift critics were quick to highlight. The timing is not accidental. Zuckerberg’s comments land alongside the policy fight, and his broader vision of a future of “personalised superintelligence,” a system tailored to each individual rather than a single, centrally controlled model built to be safe on everyone’s behalf, frames openness as the winning side of history and closed, centralised systems as the losing one.
What Washington still has to decide
Officials weighing export controls and model restrictions are balancing competitiveness against security. Zuckerberg has now placed Meta firmly on the competitiveness side of that ledger. What remains unresolved is the harder judgement: whether open-weight releases make American AI stronger by spreading capability widely, or simply make the technology easier for competitors, including those in China, to copy and build on. One Financial Times interview will not settle that question for regulators, but it has added a high-profile voice to the side arguing that openness is the strategic answer.
FAQ
What did Mark Zuckerberg say about blocking Chinese AI models?
Zuckerberg told the Financial Times that banning advanced Chinese AI from use in the United States is “not an effective solution.” He argued the US should instead identify where it is behind and address those gaps, and said the industry’s direction of travel has always been towards greater openness.
Which companies support open-weight AI models?
Meta, Microsoft, Nvidia, and Elon Musk have lined up behind open-weight AI. OpenAI and Anthropic have argued for tighter controls on the most capable systems on safety and national-security grounds.
Why is the policy debate about Chinese AI models happening now?
Chinese models have been closing in on US leaders, and Moonshot AI released Kimi K3 in mid-July. On 24 July, twenty-five organisations signed a letter titled “Open Weights and American AI Leadership,” arguing openness is the surer route to continued US primacy in AI.
Related coverage
- Trump administration reportedly reviving push to ban Chinese AI models after Kimi K3 launch
- OpenAI says its models escaped a sandbox and breached Hugging Face
This article summarizes reporting from thenextweb.com.