Nvidia CEO Jensen Huang joined X on July 24, 2026, and used his first-ever post on the platform not to introduce himself but to share an open letter arguing that US policy should support open-weight AI models alongside closed, proprietary ones.
What the letter says
The letter, signed by roughly two dozen companies including Nvidia, Microsoft, Meta, IBM, and Palantir, was addressed to Washington policymakers as they weigh restrictions on open-weight AI amid rising competition from Chinese labs. Huang wrote that “open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty,” and argued that “the world needs both frontier closed models and frontier open models.” The signatories warned against repeating what they characterized as a mistake narrowly avoided in the 1980s, when some in the software industry pushed to restrict open, freely modifiable code. Notably, OpenAI and Anthropic did not sign the letter.
The policy backdrop
The letter lands as Washington debates how to treat increasingly capable open-weight models released by Chinese labs, including Moonshot AI’s Kimi K3, a 2.8-trillion-parameter model released July 16 that closed much of the performance gap with leading US systems and has since become a flashpoint in US-China AI tensions. Some policymakers have floated tighter restrictions on open-weight releases as a national-security measure; Huang and the other signatories argue that curbing open models domestically would cede ground to those same competitors rather than contain them, since open-weight development would continue abroad regardless.
Why it matters
The split among frontier labs is itself notable: Nvidia, Microsoft, and Meta — all of which either ship open-weight models themselves or depend on a healthy open ecosystem for hardware and platform sales — are pushing back against restrictions that closed-model specialists like OpenAI and Anthropic have been comparatively quiet on or more cautious about. Huang choosing this argument, rather than a product announcement, for his first public post on X underscores how central the open-versus-closed model debate has become to Nvidia’s own commercial interests, since a broader open-weight ecosystem drives demand for the chips it sells to a much wider set of AI developers than a handful of closed frontier labs would.