The impressive capabilities of Chinese lab Moonshot’s Kimi K3, the largest open-weight large language model, have started a debate that combines two things: the economic possibilities of American AI giants and the future of LLMs as a technology.

OpenAI’s head of strategic futures, Dean W. Ball, went so far as to argue that the US government should find a pretext to create fear, uncertainty, and regulatory mistrust around new models, since open-weight models must necessarily deter capital spending by frontier labs.

People freaked out, and tech luminaries like Yann LeCun and Martin Casado argued that open software can accelerate innovation and coexist with proprietary projects. Ball soon walked back his claims that a regulatory offensive was the White House’s “best strategy” and that open-weight models necessarily slow advances in the technology.

However, Axios reports that the Trump administration is considering banning the K3 and other advanced Chinese models at the behest of US border labs. Another Politico report said the Commerce Department would not take that step anytime soon.

The benefit for leading AI companies is clear: open models, running on standalone infrastructure or within large enterprises, offer cheaper intelligence than class-leading models from Anthropic or OpenAI. If users increasingly spend more outside of closed labs, that means a lower return on their huge investments in model training.

That vision extends far beyond OpenAI. “Strong, frontier-caliber open source models will squeeze margins and drive down prices for frontier companies,” Braden Hancock, co-founder of Snorkel AI and research partner at the Laude Institute, told TechCrunch. “It won’t necessarily mean that the amount of AI usage goes down a little bit. You know, obviously, quite the opposite.”

That’s not a problem for people who don’t own shares in Anthropic and OpenAI. AI will continue to proliferate. So what is the justification for the government to prevent Americans from buying anything in our seemingly free markets?

Concerns about Chinese models are of various colors. One is to protect American data from the Chinese government; The United States has banned the import of modern Chinese electric vehicles over concerns about their data collection. But experts tend to think that open weight models running on American servers are unlikely to leak data to China, although it is not impossible that such a thing could be done.

Another is that models may have an implicit bias toward CPR, but it’s unclear what that might mean, for example, for encoding tasks.

A third common concern is that the Chinese models lack the barriers that the US government has imposed (through an opaque process), which are intended to prevent major US LLMs from being used to exploit closed computer systems or create weapons. However, those same barriers can make American companies more vulnerable: David Sacks, the venture capitalist and Trump adviser, has been sharing cases of American companies turning to Chinese LLMs to close security gaps when American border models refuse to perform the tasks.

But the most important motivation for restricting the models is fear that China could overtake the United States if border labs slow down.

Sam Bresnick, a China researcher at Georgetown’s Center for Security and Emerging Technology, says the growing importance of AI to U.S. military operations gives the United States a reason to support continued investment in AI at border labs. But the whole question, he says, is complicated.

“Why should the weight of the US government be aimed at protecting these companies from competitors who are being excluded from the US market based on their origins?” Bresnick asks.

Open AI advocates say frontier companies are creating a false binary between innovation and closed models.

“The biggest impact of these open source models coming from China is not so much that they sneak in through back doors, but rather that they own the innovation,” Hancock told TechCrunch. “Effectively, you end up with an expanded workforce in your model. PyTorch became the industry standard because it was open source, so the entire community could contribute instead of just one company, and it grew and grew, and the rest of the deep learning libraries died in comparison.”

Hancock and other advocates fear that Chinese LLMs will become the focus of international research. U.S. graduate programs already rely largely on open Chinese models, and Hancock says half of the work students study comes from Chinese institutions, and U.S. frontier labs are increasingly reluctant to share their work widely.

“Restricting open models would not make AI safer,” said Clem Delangue, CEO of Hugging Face, a platform for open AI collaboration. “It would simply obscure the risks, concentrate power in the hands of a few, and make it harder for the next generation of builders, researchers, academics, nonprofits, and governments to participate in making AI safer and more beneficial for everyone.”

Bresnick says the real way to stop China would be to focus more on chip export controls. A better way to preserve American leadership in AI would be to stop selling Nvidia H200 processors to China. “That,” he says, “could potentially keep us out of this thorny debate over banning open source technologies that a large number of American companies want to use.”

Part of the problem is the uncertainty around the economics of AI. “Neither the open business model nor the proprietary business model is solved. AI companies are struggling to figure out how to make money with their tools, especially as training costs have to get higher and higher,” Bresnick says.

The same challenges present in the US are also present in China, where AI companies are also struggling to generate revenue and access computing power, and the government is seen as encouraging open releases for political reasons despite the challenge of capitalizing on them.

Some American companies, including Thinking Machines Lab and Nvidia, are trying to do business by launching open models. Hancock notes that Nvidia would be better off “if there were dozens or hundreds of companies building AI rather than two or three that are well-capitalized enough to make their own chips,” which is one of the reasons behind its investment in Nemotron, a collection of open models.

“The main point is that the United States would really benefit from having its own very capable and much less expensive open models,” Bresnick said. “It just clashes with the approach that frontier labs have taken.”

With additional reporting by Rebecca Bellan.