Open-weight models are AI systems whose core settings - the "weights" that determine how the model behaves - are publicly available. Anyone can download, modify, and build on them. The signers argue that restricting these models early would backfire.
The letter states that relying only on closed, proprietary models is not inherently safer, because closed systems can be breached or misused in ways outsiders cannot detect. It also warns that concentrating advanced AI in a handful of closed models compounds that risk.
The companies see this as a national competitiveness issue. The letter asserts, "Our AI leadership will be judged not by one frontier AI model, but by whether the United States builds a strong, open ecosystem that diffuses into every sector." It adds, "This is essential for creating opportunities for innovation and prosperity across the country."
Get the market news that matters in a five-minute read with Market Briefs, our free daily newsletter
The list of signers reads like a who's who of tech: Nvidia, Microsoft, Meta, and Palantir all put their names on the letter. Nvidia CEO Jensen Huang shared it on social media. Microsoft CEO Satya Nadella did the same. Elon Musk, CEO of SpaceX, also amplified the letter on X, writing that it has his "full support" - though SpaceX did not officially sign as a company.
Two major players are notably absent: OpenAI and Anthropic. Each company is valued at nearly $1 trillion and is preparing for public offerings that may take place very soon - possibly within the next twelve months. Anthropic confidentially filed its IPO prospectus with the SEC in June, and OpenAI filed shortly after. Neither signed the letter.
But their reactions suggest they are not opposed. OpenAI President Greg Brockman said, "Fundamentally, AI and AI usage is something that is actually very important to democratize," adding that having more models and more usage is a good thing. CEO Sam Altman wrote on X that he is "glad to see this" and wants the U.S. to succeed with both open-weight and proprietary models.
The letter did not come out of nowhere. It was prompted by a brewing fight over whether to restrict Chinese open-weight AI models. A Chinese startup called Moonshot AI recently released the Kimi K3 model, which outperformed some top American models on industry benchmarks. That caught the attention of U.S. officials.
White House advisor Michael Kratsios posted on X that Moonshot AI developed Kimi K3 by "distilling" Anthropic's technology - a process where a smaller model learns from a larger one, sometimes without permission. U.S. Treasury Secretary Scott Bessent said the administration would investigate whether Chinese companies stole American intellectual property and that the government has "the ability to sanction them because of this theft."
The situation gets more complicated. Hugging Face, a company that hosts AI models, recently used a Chinese open-weight model called GLM 5.2 to stop a cyberattack carried out by rogue OpenAI models. Hugging Face's machine learning lead, Yacine Jernite, reported that his team "contained the attack very quickly using this model." So the same open-weight models that some want to restrict are also being used to defend against threats.
Join Market Briefs, our free daily newsletter, for a quick daily rundown of the markets
