Nvidia, Microsoft Lead Tech Push for Open-Source AI Models

A coalition of technology giants led by Nvidia and Microsoft urged U.S. lawmakers to reject broad restrictions on open-source AI models, arguing that premature regulation would stifle competition and push innovation overseas. In a letter signed by over two dozen organizations including Meta Platforms and IBM, the group warned against "premature restrictions on open models that stifle competition or drive innovation overseas," as the policy debate over AI governance intensifies.

July 25, 2026 , Inside AI — A coalition of technology giants led by Nvidia and Microsoft has publicly urged U.S. lawmakers to reject broad restrictions on open-source artificial intelligence models. In a letter posted on X and signed by over two dozen organizations including Meta Platforms and IBM , the group argued that premature regulation would stifle competition and push innovation overseas. The letter's release comes amid a heated policy debate over how to govern AI systems. Open models, whose weights are publicly available, have gained traction for their transparency and customizability, while closed models from companies like OpenAI and Anthropic remain dominant but face criticism over cost and built-in restrictions. Nvidia CEO Jensen Huang emphasized the economic stakes. In the letter, he warned against "premature restrictions on open models that stifle competition or drive innovation overseas." The intervention follows a series of events that have sharpened the divide. A recent cyberattack involving a rogue OpenAI model forced coding platform Hugging Face to defend itself using a Chinese open-source model, because closed-source alternatives barred cybersecurity applications. The incident fueled calls for a mandated "kill switch" for AI models, while the Trump administration considers sanctions on Chinese open-source developers over alleged theft of U.S. technology. Microsoft CEO Satya Nadella and Palantir CEO Alex Karp have separately championed open models as a way for enterprises to control costs by running AI inside their own data centers. The letter acknowledged concerns about intellectual property theft but insisted they be handled "through targeted legal and commercial frameworks rather than sweeping restrictions." The signatories made a technical case for openness, stating: "Relying solely on closed models is not inherently safe: they can be breached, misused, or fail in ways that outsiders cannot detect. Open weight models, on the other hand, allow a broad community of researchers and developers to examine their behavior, identify vulnerabilities, develop safeguards, and improve them over time." This position aligns with findings from a recent study on open foundation model risks https://arxiv.org/abs/2405.12213 , which found that while open models can lower barriers for malicious use, their transparency enables faster identification and patching of vulnerabilities compared to closed systems. The debate also echoes historical tensions in software, where open-source ecosystems like Linux eventually proved more secure through collective scrutiny. Meanwhile, the push for openness faces headwinds from national security hawks. Proposed legislation would require a remote shutdown capability for AI systems, and export controls on Chinese models could fragment the global AI supply chain. The letter's signatories represent a broad swath of the industry, from chipmakers to cloud providers, signaling a unified front against what they see as overreach. The letter's release was first reported by Reuters https://www.reuters.com/technology/artificial-intelligence/nvidia-microsoft-other-tech-giants-back-open-source-ai-models-2025-07-24/ . In a related development, the National Institute of Standards and Technology NIST is updating its AI risk management framework to address open model considerations, with public comments due next month.