Jensen Huang’s open AI model push 

Nvidia founder and CEO Jensen Huang made his debut post on X (formerly Twitter) on Friday, using the moment to back open-weight artificial intelligence models over concentrating development in closed systems, as Washington weighs new restrictions on foreign open models. 

Huang shared a letter that Nvidia and 24 other companies, including Microsoft, Meta, and Hugging Face, signed opposing limits on open models.

Why Huang chose this for his first post

Huang, who had stayed off the platform even as rival chipmakers’ CEOs built a presence there for years, used his opening post to make a policy argument rather than a product announcement. He wrote that AI will transform every industry, power every company, and be built by every country, and that open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.

The letter argues that AI will reshape every industry and economy, and that the US should foster both frontier proprietary models and open-weight models to maximize innovation, competition and resilience. 

Open-weight models allow users to download, inspect, modify and operate the underlying model on their own infrastructure, which the letter says can reduce dependence on individual cloud or model providers.

The China angle

The post lands at a sensitive moment for US AI policy. About a year ago, the government’s own AI Action Plan backed open models and called them a strategic asset for the country. Officials are now weighing curbs on Kimi K3 and other Chinese models. Kimi K3, a 2.8-trillion-parameter model from China’s Moonshot AI, launched on July 16 and ranks among the best models anywhere, open or closed.

The concern isn’t new for Nvidia. In January 2025, a low-cost model from China’s DeepSeek sent shockwaves through the company, wiping nearly $600 billion off its value in a single day — a record at the time.

A safety argument, too

Beyond competitiveness, the signatories made a security case for keeping models open. The letter argues that open models let many teams check the underlying code, so flaws can be spotted and fixed quickly.

Huang’s post signals that Nvidia — whose business depends on AI development staying broad-based rather than concentrated among a handful of closed-model providers — is actively lobbying against tighter restrictions on open-weight systems, even as US policymakers grow wary of China’s rapid progress in the space. Whether the letter shifts the administration’s thinking on Kimi K3 and similar models remains to be seen.

Global tech leaders back the open-model push

The letter drew swift, public endorsement from some of the industry’s most prominent figures. Microsoft chairman and CEO Satya Nadella called open-weight models “essential to a healthy AI ecosystem.” 

Meta founder and CEO Mark Zuckerberg backed the move too, writing that he was “proud to support this,” and framed open source as a force against industry centralisation. 

OpenAI CEO Sam Altman said he was “glad to see this” effort toward US leadership in both open and proprietary AI. Microsoft vice chair and president Brad Smith said open-weight models “play a critical role in building a broad AI ecosystem.” 

Dell Technologies founder and CEO Michael Dell argued that “America’s tech leadership has always come from ecosystems where the best ideas compete.” 

Former OpenAI CTO Mira Murati added that for AI to draw on the world’s distributed expertise, “it must itself be distributed.