Every once in a while, a Chinese company releases a new AI model and Americans freak out.

That’s what happened last week when China’s Moonshot AI unveiled the Kimi K3, which surpassed a number of notable benchmarks. By most accounts, the Kimi K3 rivals some leading US models for a fraction of the cost.

The new model triggered another round of panic that China is closing in on the United States in the race to corner the AI ​​market and accusations that the Chinese are training their models based on work already done by Anthropic, OpenAI and Google. (Earlier this month, Business Insider’s Ali Barr highlighted the irony of that accusation.)

Increasingly, the tension boils down to a clear strategic divide between the two countries: the Chinese have adopted open source or open weight models, while the United States remains largely closed.

Debate on open or closed models

A debate broke out on X over the weekend after an OpenAI executive posted a lengthy reaction to Kimi K3.

“Personally, I am surprised that the Chinese state continues to allow open access to such good models, given the potential risks,” Dean Ball, former senior AI advisor to President Donald Trump and recently chief strategist at OpenAI, wrote in X.

He said the open weight strategy would lead to outright “AI communism” and that open models can be “decelerationist” because they “deter AI capital spending.”

However, what really drove the conversation was this:

“My guess is that the Trump Administration will at some point realize that its best strategy here would be to create large amounts of regulatory risk around the use of Chinese open-weight models,” he wrote.

Ball suggested that manufacturing fear, uncertainty, and doubt (or “FUD,” for those in the know) into the regulatory process would cause most U.S. companies to avoid using open models.

He later clarified that this was his prediction, not a recommendation, and that he supports open source to the point where AI becomes too dangerous, which he said would be a “sad day.”

The response was quick and widespread.

Manufacturing regulatory confusion in support of American AI labs sounds a lot like “regulatory capture.” That’s when government agencies tasked with regulating an industry design rules to support it, often following the advice of the industry’s own experts.

Anthropic and OpenAI have argued that their models are too powerful to be made open and that doing so would be dangerous and would allow anyone to use their tools for any purpose, with little oversight. A closed system gives the manufacturer greater overall control, including over security, access and pricing. Laboratories have warned that open weight models leaving China are a threat to national security and their businesses.

David Sacks, a venture capitalist who was Trump’s first AI and cryptocurrency czar before moving in March to co-chair the president’s Council of Science and Technology Advisors, said the “weaponization of regulatory uncertainty” was “completely unacceptable.”

“We are at a critical inflection point in AI policy. The major closed labs, already a duopoly in terms of revenue from AI models, want the government to eliminate their open source competition,” he wrote on X in response to Ball’s post. That “duopoly” refers to OpenAI and Anthropic. “They’ve put their cards on the table. It’s time for the rest of Silicon Valley – the vast majority that still values ​​open competition – to do the same.”

Sacks’ “All-In” podcast co-host and fellow VC, Chamath Palihapitiya, was equally blunt: “The future is open source,” he wrote on X. “We have to accept it and move on.”

Suhail Doshi, a prominent software engineer and entrepreneur, said American AI labs trained their products with “humanity’s data and didn’t pay a cent.”

“Any lobbying or legislation calling for banning open weight models in the name of ‘distillation’ is total nonsense,” he wrote on X. “This is a fight against future American innovation.”

A Citrini Research analyst who goes by the name Jukan at X disagreed with Ball and the warnings about a Chinese acquisition, writing that open source models do not automatically position companies to dominate. He said DeepSeek, for example, can operate more efficiently, keeping token costs lower, due to its proprietary operations, not just its open source framework.

“Chinese companies may lack enough computing power to meet all inference demand on their own, but they are not selling at a loss or defaulting on their training costs,” Jukan wrote.