A coalition of more than two dozen technology companies has called on U.S. policymakers to avoid imposing what they describe as premature restrictions on open-weight artificial intelligence models, arguing that open AI development is essential to maintaining innovation, competition, and American leadership in the rapidly evolving AI industry.

The appeal comes as policymakers and industry leaders debate how to respond to increasingly capable open-weight AI models from China, raising broader questions about balancing national security, intellectual property protection, and technological competitiveness.

Technology Companies Advocate for an Open AI Ecosystem

Nvidia, Microsoft, Meta, Palantir, and over 20 additional technology firms jointly signed a public letter encouraging regulators to preserve access to open-weight AI models, which allow developers to download, modify, and deploy models on their own infrastructure.

According to the companies, restricting these models too early could reduce competition, limit innovation, and encourage AI development to move outside the United States. The group argued that open-weight models help distribute AI capabilities more broadly instead of concentrating them among a handful of providers operating proprietary systems.

The signatories also challenged the assumption that closed AI models inherently provide stronger security, noting that proprietary systems can still be vulnerable to misuse or failures that remain difficult for outside researchers to identify.

Nvidia Chief Executive Jensen Huang and Microsoft Chief Executive Satya Nadella publicly shared the letter, underscoring the importance many large technology companies place on maintaining an open AI ecosystem.

Chinese AI Advances Intensify Policy Debate

The industry’s appeal comes as Chinese developers continue to make significant progress in open-weight AI.

Recent attention has focused on Moonshot AI’s Kimi K3 model, which has demonstrated competitive performance against several leading U.S. AI systems on selected industry benchmarks. The model’s release has intensified discussions in Washington regarding the competitive landscape between American and Chinese AI developers.

Government officials have also raised concerns about potential intellectual property misuse. U.S. officials have indicated they are examining whether some Chinese AI companies may have benefited from unauthorized use of American technology, including through AI model distillation, a training technique that leverages outputs from more advanced models.

While acknowledging those concerns, the coalition argued that any cases involving unlawful technology transfer should be addressed through targeted legal and commercial measures rather than broad restrictions that could slow legitimate AI research and development.

Industry Divisions Reflect Different AI Strategies

Notably absent from the letter were OpenAI and Anthropic, two leading developers of proprietary AI models that are reportedly preparing for potential public offerings.

Although the companies did not join the initiative, OpenAI President Greg Brockman recently expressed support for broad access to AI technologies, emphasizing the importance of expanding AI availability while noting he had not participated in discussions regarding potential bans on Chinese open-weight models.

The differing positions highlight the growing divide within the AI industry between companies emphasizing open-weight development and those focused primarily on proprietary foundation models.

Looking ahead, policymakers are expected to continue weighing the benefits of open AI innovation against national security and intellectual property concerns. The outcome of these discussions could significantly influence the future direction of AI development, international competitiveness, and regulatory policy as governments seek to encourage innovation while protecting critical technologies.