Nvidia, Palantir, and Meta Warned Trump’s White House Against Restricting Open-Weight AI — Here’s Why

What the letter said

The coalition — which also included Microsoft, IBM, Hugging Face, Mistral, Dell, Andreessen Horowitz, Y Combinator, and the Linux Foundation — argued that open-weight AI has been a critical engine of American AI leadership and that restricting it would cede ground to rivals like China . Open-weight models are systems whose trained parameters can be downloaded, inspected, modified, and run on users' own infrastructure .

Nvidia CEO Jensen Huang broke his silence to publish the letter as his first-ever post on X, arguing that "the world needs both frontier closed models and frontier open models" .

Key concerns raised by the signatories

• Restricting open-weight models could backfire on national security. Palantir and other defense-focused signatories warned that broad limits could weaken the U.S. AI ecosystem that supports both commercial innovation and national security applications .
• Openness is itself a path to safety. The letter argued that transparency and broad community scrutiny of open-weight models enable faster identification of vulnerabilities and better alignment research, making openness "one of the most important paths to AI safety" . The signatories also wrote: "Relying solely on closed models is not inherently safe: they can be breached, misused, or fail in ways that outsiders cannot detect" .
• Premature regulation would stifle competition. Rather than restricting open models, the companies urged the government to focus on expanded public compute access, shared infrastructure, and clear safety guidelines — not blanket prohibitions .

Notably absent from the list of signatories were OpenAI and Anthropic, two prominent closed-model companies .

The US-China AI competition context

The letter was published as Washington was actively debating how to respond to allegations that Chinese AI labs were stealing U.S. intellectual property and using open-weight models to rapidly close the gap with American frontier systems . Some Trump administration officials had signaled they might take action against open-weight distribution precisely because Chinese firms were benefiting from it . Reports indicated the administration was considering banning Chinese open-weight models and potentially issuing sanctions against Moonshot AI, the creator of the Kimi K3, described as the world's most powerful open-weight AI model .

The signatories pushed back on that logic, arguing that restricting open models would push global developers — including allies and neutral actors — toward Chinese alternatives (such as DeepSeek's open models), strengthening China's AI ecosystem rather than weakening it . Their core argument was that the U.S. should win the AI competition through faster innovation and broader access, not through containment .

"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 signatories wrote .