Tech Giants Urge US to Spare Open-Weight AI Models From Restrictions

A coalition of 25 technology companies released an open letter July 24 urging the Trump administration to avoid imposing broad restrictions on open-weight AI models, warning that premature rules would undermine US competitiveness as Washington weighs responses to Chinese AI advances.

The letter, published on Microsoft's corporate responsibility page under the title "Open Weights and American AI Leadership," was signed by Nvidia, Microsoft, Meta, Mistral, Hugging Face, Palantir, Andreessen Horowitz, Y Combinator, Mozilla, The Linux Foundation, and 15 other organizations. Notably absent were OpenAI, Anthropic, Google DeepMind, and SpaceX—companies whose closed-model business models could benefit from restrictions on open-weight competitors.

The industry mobilization comes as the White House accuses Chinese firm Moonshot AI of distilling Anthropic's Fable model to develop Kimi K3, released in July 2026. Michael Kratsios, director of the White House Office of Science and Technology Policy, alleged Moonshot built a sophisticated internal platform for large-scale distillation against US models. Kratsios also alleged Moonshot accessed banned Nvidia GB300 chips through Thailand, though he offered no public evidence for this claim.

Treasury Secretary Scott Bessent warned July 21-22 that the US could sanction Chinese AI companies found to have stolen intellectual property through distillation. The industry letter pushes back, distinguishing between lawful distillation—a widely used technique for model improvement—and unlawful extraction of value from closed models.

"Unlawful efforts to extract value from closed models raise legitimate concerns," the letter states. "Those concerns should be addressed through targeted legal and commercial frameworks rather than sweeping restrictions on techniques that play an important role in AI innovation."

Nvidia CEO Jensen Huang posted on X—his first-ever post on the platform—that "open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty." Elon Musk replied expressing full support.

Replit CEO Amjad Masad told TechCrunch that "banning Chinese open models is as good as banning open models in general," noting that Thinking Machines Lab's Inkling model was trained with help from Moonshot's Kimi 2.5. The interconnected nature of the ecosystem was underscored when Hugging Face was unable to defend against an AI-powered breach using commercial frontier models because their guardrails blocked defensive efforts, forcing the company to use Chinese firm Z.ai's GLM 5.2 open-weight model instead.

The letter frames open-weight models as critical to US competitiveness, drawing parallels to the 1980s open-source software movement. Signatories argue that infrastructure providers like Nvidia and cloud companies benefit from commoditized models that drive GPU sales and cloud capacity usage.

"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," the letter states.

The administration's response will signal whether US policy prioritizes protecting incumbent AI companies or fostering a broader, more competitive ecosystem as Chinese open-weight models rapidly close capability gaps despite having fewer computational resources.