Nvidia, microsoft, meta, Palantir On Friday, more than 20 other companies released a letter urging policymakers to avoid “premature restrictions” on promiscuous artificial intelligence models that would “hinder competition or promote innovation abroad.”
Openweight AI models, which users can download, modify, and run on their own infrastructure, have been the subject of intense debate within the tech industry in recent weeks.
China’s open weight model is gaining traction against major products from US companies. OpenAI and humanmainly develops its own closed model. Officials and executives are considering whether to restrict access to Chinese models in the US
Chinese startup Moonshot AI compounded concerns earlier this month when it released a model called Kimi K3 that outperformed state-of-the-art U.S. products on some industry benchmarks. U.S. Treasury Secretary Scott Bessent told CNBC on Tuesday that the Trump administration will investigate whether Chinese companies are stealing American intellectual property, adding that the government “has the ability to impose sanctions on Chinese companies for this theft.”
But a group of US tech companies warned against rash action in a letter on Friday. They write that an open-weight model strengthens competition and ensures that the benefits of technology are “shared widely, rather than concentrated in the hands of a few.”
“Relying solely on a closed model is inherently insecure; it can be compromised, exploited, or fail in ways that cannot be detected by outsiders,” the letter said. “And concentrating advanced AI capabilities behind a small number of closed models further increases that risk.”
Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella both shared the letter on their personal social media accounts.
Elon Musk runs an AI business at a rocket company. space xalso applied the letter on social media, writing in a post to X that they “fully support it.” SpaceX did not formally sign the letter.
OpenAI and Anthropic did not sign the letter. Both companies, each valued at nearly $1 trillion, are preparing for major initial public offerings that could occur as early as this year. Anthropic confidentially filed a prospectus with the Securities and Exchange Commission in June, and OpenAI followed suit a few days later.
OpenAI President Greg Brockman said Thursday that the company believes in broad access and is not involved in any conversations with the Trump administration about a potential ban on Chinese-made open-grade models in the United States.
“Fundamentally, I think AI and the use of AI is very important for democratization,” Brockman told reporters at a news conference in New York City. “So, for me, on some kind of deep level, I think having more models and having more uses is a good thing.”
OpenAI CEO Sam Altman referenced the letter in a post on X on Friday, saying he wants the U.S. to win with both the open-class and proprietary models and is “glad to see this.”
Earlier this month, AI company Hugging Face used Chinese company Z.ai’s open weight model to thwart a cyberattack carried out by a rogue OpenAI model. OpenAI disclosed the attack on Tuesday, characterizing it as an “unprecedented cyber incident.”
OpenAI CEO Sam Altman speaks to reporters after a meeting at the Capitol in Washington, DC, USA on June 3, 2026.
Kylie Cooper | Reuters
Yacine Jernite, head of machine learning at Hugging Face, told CNBC that the company initially tried to use Anthropic’s Fable 5 to analyze the attack, but it was unsuccessful because the model’s guardrails couldn’t determine whether Hugging Face was trying to protect itself.
Jarnight said Hugging Face utilized Z.ai’s model GLM 5.2 and was able to contain the attack “very quickly using this model.”
White House adviser Michael Kratsios said Wednesday that China’s Moonshot AI extracted Anthropic’s technology to develop the Kimi K3 model. Distillation is a term used to describe an AI training method that uses the output from an existing, more powerful model to build a smaller, less powerful model.
In his post on X, Kratsios wrote that legal AI distillation plays an important role in the open innovation ecosystem, but warned that “large-scale clandestine industrial distillation aimed at stealing America’s proprietary technology” is “unacceptable.”
In a letter Friday, the U.S. tech companies said concerns about illegal distillation should be addressed through “targeted legal and commercial frameworks” rather than “blanket restrictions on technologies that play a key role in AI innovation.”
“Our nation’s AI leadership will not be judged by one frontier AI model, but by whether America builds a strong, open ecosystem that permeates all sectors,” the letter states. “This is essential to creating opportunities for innovation and prosperity across the country.”
clock: TrustedSec’s David Kennedy says AI is forcing the cyber industry to rethink how it operates.
