Key Takeaways
- Nvidia, Microsoft, Meta, Palantir and more than 20 other companies urged policymakers not to impose “premature restrictions” on open-weight AI models.
- The letter says open-weight systems support competition and broader access, while arguing that restrictions could slow innovation or push it overseas.
- The debate comes as Chinese open-weight models gain attention and U.S. policymakers consider how to respond.
What happened
Nvidia, Microsoft, Meta, Palantir and more than 20 other companies released a letter on Friday urging U.S. policymakers to avoid premature limits on open-weight artificial intelligence models.
In the letter, the companies warned that broad restrictions could “stifle competition or drive innovation overseas.” They argued that open-weight models help spread the benefits of AI more broadly rather than concentrating them among a few large firms.
Open-weight models are downloadable systems that users can modify and run on their own infrastructure. That has made them a major point of debate inside the AI industry, especially as policymakers consider whether Chinese models should face restrictions in the United States.
The letter landed amid growing attention on Chinese open-weight systems. CNBC reported that Moonshot AI, a Chinese startup, recently drew notice for a model called Kimi K3 that outperformed leading American offerings on some industry benchmarks.
The companies behind the letter also argued that closed models are not automatically safer. The letter said that closed systems can still be breached, misused or fail in ways outsiders cannot see, and that concentrating advanced AI behind a small number of closed models adds to that risk.
Nvidia CEO Jensen Huang and Microsoft CEO Satya Nadella both shared the letter on their personal social media accounts. Elon Musk also amplified it on X, saying it had his “full support,” though SpaceX did not officially sign the letter.
OpenAI and Anthropic did not sign the letter. CNBC noted that both companies are preparing for possible public listings, with confidential filings made this year.
Why it matters
The letter reflects a broader fight over how the U.S. should manage AI competition and security at the same time.
On one side is the argument that open-weight models are essential to a healthy AI ecosystem because they let more developers, companies and researchers build on top of existing systems. On the other is the concern that wide availability could make it easier for rivals or malicious actors to reuse advanced model capabilities.
The companies signing the letter framed the issue as one of national competitiveness. They said U.S. leadership should be measured not by a single frontier model, but by whether the country can build an open ecosystem that spreads into many sectors.
The timing also matters because the policy discussion is overlapping with fears about intellectual property theft and model distillation. U.S. Treasury Secretary Scott Bessent said this week that the Trump administration would look into whether Chinese companies were stealing American intellectual property. Separately, White House advisor Michael Kratsios said legitimate distillation can support open innovation, but that large-scale covert distillation aimed at stealing proprietary U.S. technology is unacceptable.
The letter pushes back on the idea that policymakers should respond to those concerns with sweeping limits on the underlying model format. Instead, the companies said worries about unlawful distillation should be handled through targeted legal and commercial frameworks.
What to watch
The biggest question is whether U.S. policymakers treat open-weight models as a distinct category worth lighter-touch rules, or whether security and trade concerns lead to broader restrictions.
It is also worth watching how the divide inside the AI industry develops. OpenAI and Anthropic did not sign the letter, even though their leaders have separately said they support broad access to AI in principle.
Another thread to follow is whether Chinese open-weight models continue to gain credibility in U.S. technical discussions. CNBC reported that Hugging Face recently used an open-weight model from the Chinese company Z.ai to help contain a cyberattack, after an attempt to use an Anthropic model did not work for that use case.
For now, the letter shows that some of the biggest names in U.S. tech want policymakers to separate concerns about misuse from the broader case for open-weight AI development.
