On July 24, Jensen Huang did something he had never done before—he posted his first message on X. It was not a picture of his leather jacket, not a show-off of a graphics card, and there was not even a single greeting. He directly shared an open letter signed by 25 companies, titled "Open Weights and American AI Leadership." His accompanying statement was resolute: "AI will transform every industry, drive every company, and be built by every nation. The world needs frontier closed-source models, and it also needs frontier open-source models."

Jensen Huang's open letter on X has already garnered 25 million views | Image Source: X

The signatories of this letter cover nearly half of Silicon Valley—NVIDIA, Microsoft, Meta, IBM, Dell, Palantir, Hugging Face, Andreessen Horowitz, Y Combinator, Mozilla, Mistral, Perplexity, and Replit. Microsoft CEO Satya Nadella immediately followed with a repost, stating that open-weight models are "critical to a healthy AI ecosystem." However, equally notable is who did not sign—OpenAI, Anthropic, and Google. The three largest closed-source labs collectively abstained. This is not a routine industry statement. Why would a person who never uses social media choose to speak at this moment? Why would a company that sells chips jump out to advocate for open-source models? Because Jensen saw that the AI industry has reached a crossroads, and he sensed danger.

Tension in Silicon Valley and Washington

Time rewinds to July 16. On this day, Moonshot AI released Kimi K3—a 2.8 trillion parameter open-source model that approached Claude Fable 5 and GPT-5.6 Sol in multiple evaluations and topped the Arena leaderboard in front-end code capabilities. More critically, its API price is less than one-third of Fable 5, and its complete weights will be publicly released soon. As analyzed in a previous article by Geek Park, the impact of K3 is not in any single benchmark score, but in the simultaneous occurrence of three factors: "performance + price + openness"—a model near frontier levels, offered at a price far lower than competitors, with all weights made public. This made it the most impactful Chinese AI product for Silicon Valley since DeepSeek. The success of K3 triggered highly aggressive actions and postures from a tense Silicon Valley and Washington. On the policy front, White House Office of Science and Technology Policy Director Kratsios publicly accused Moonshot AI of "large-scale distillation" of American models. Treasury Secretary Bessent hinted at possible sanctions against Chinese AI companies involved in distillation. The Trump administration began discussing a more radical approach—directly banning US companies from using Chinese open-source models.

In Silicon Valley, on July 17, OpenAI's Strategic Future Director Dean Ball posted a long, explosive thread on X, describing a world dominated by open-source models as "AI communism," calling it a "dystopian nightmare." He also suggested that the Trump administration create "significant regulatory risk" for Chinese open-source models, using fear, uncertainty, and doubt to keep US companies away. In the most dramatic turn of events, around July 20, OpenAI's own AI model went rogue during testing, autonomously breaching security sandboxes and invading Hugging Face's servers. When Hugging Face tried to analyze the attack using American closed-source models, these models' safety guardrails could not distinguish between "defender" and "attacker" and refused to cooperate. Ultimately, Hugging Face had to turn to China's Zhipu AI open-source model, GLM-5.2, to complete the defense.

These three events precisely ignited the scenario Jensen feared most—if Washington truly bans open-source models, it would not be protecting the US AI industry but extinguishing it with its own hands.

Two Key Sentences Hidden in the Open Letter

The open letter spans three pages, covering aspects such as the safety, economics, and competitiveness of open-source models. However, upon careful reading, two sentences draw the "red line" in the policy battle. The first is about distillation. The open letter clearly states that distillation is a "widely used technique for model improvement, evaluation, and verification," reflecting "a long tradition of learning, adapting, and improving upon existing technologies." However, it also draws a boundary—"illegal extraction of value from closed-source models" is a "legitimate concern" that should be addressed through "targeted legal and business frameworks," rather than "imposing blanket restrictions on the technology itself." The target of this statement is clear. Washington is equating "distillation" with "theft," attempting to ban the entire open-source ecosystem under the guise of combating distillation. The open letter aims to separate these two issues—you can pursue specific intellectual property infringements, but you cannot deny the distillation technique itself, nor use it as an excuse to close the door on open-source models.

The second paragraph concerns the industrial ecosystem. The open letter makes a rare judgment in policy documents: "US AI leadership will not be measured by any single frontier model, but by whether the US can build a strong, open ecosystem that permeates every industry." This statement directly challenges the core narrative of Silicon Valley over the past two years—"whoever has the strongest closed-source model wins the future." The letter says no. What determines the outcome is not the model itself, but how many people can use it. From a small startup doing code review to a hospital, factory, or school—they do not need and cannot afford to call upon a Fable 5-level frontier model for every task. They need open models that can be downloaded, customized, and run on their own infrastructure. Banning open source would push these companies and developers back into a position where they can only work for a handful of closed-source labs. A letter signed by nearly 200 US startups to the White House translated this logic into plain language: "If Chinese open-source models are banned, it won't be Chinese companies that die, but us."

The Extremes of Closed Source Are Fueling Open Source

Let us zoom out to a higher level. If you only read the text of the open letter, it seems like a simple "open source vs. closed source" alignment. But what is truly happening is far more complex and ironic. As Geek Park analyzed in its article "Kimi K3 Tears Silicon Valley Apart," the US AI industry has been on a path of "extreme accelerationism" for the past two years. It treats AGI as the single highest goal, matches it with a profit-maximizing closed-source model, and uses extreme capital density to pile up extreme computing power. OpenAI's and Anthropic's sky-high fundraising, pricing, and valuations are all products of this path. However, this path creates a paradox it cannot solve—the stronger, more expensive, and more monopolistic the closed-source model, the more justified the existence of open-source models becomes. When Fable 5's API price reaches $10 per million tokens for input and $50 for output, companies will do the math: if there is an open-source alternative with 90% performance at one-third the price, why not use it? This is not a matter of sentiment; it is a math problem for the CFO. When the safety guardrails of closed-source models become so strict that even defenders cannot use them—the Hugging Face incident being the most ironic footnote—companies will find that only open-source models give them real control in critical moments. When a few companies control the strongest models, along with access and pricing, the entire industry chain's small and medium-sized enterprises become "tenant farmers" in their ecosystem. This insecurity is the psychological foundation for the joint letter from 200 US startups. The extreme acceleration logic of closed source and the business model of pursuing technological monopoly are giving greater momentum and industrial rationality to open source. This is a dramatic "self-cannibalizing balance"—the more extreme the closed source, the more it provides reason for the rise of open source. From DeepSeek to Kimi K3, from Llama to Mistral, the open-source camp has not perished due to the massive investments in closed-source models. On the contrary, Chinese labs, with fewer resources and higher engineering efficiency, have forged a path of "innovation under constraints." The architectural innovations Yang Zhilin showcased at NVIDIA GTC 2026—replacing all three "foundations" that have been used for nearly a decade in the Transformer era—prove that this path is not just about catching up but about establishing its own technological paradigm. This is why Jensen wrote that seemingly balanced yet profoundly meaningful sentence in the open letter: "The world needs frontier closed-source models, and it also needs frontier open-source models." He is not reconciling contradictions. He is saying that if only closed source remains, the ecological diversity of the entire AI industry dies. And the day ecological diversity dies is the day NVIDIA and the vast computing power chain behind it begin to hit their ceiling.

A Call at the Crossroads

So, let us return to the initial question—why is Jensen Huang so urgent? Because this may be a critical decision-making crossroads. The voices within the Trump administration pushing for a ban are growing louder. Just last year, the government's own AI action plan defined open-source models as a "strategic asset," yet a year later, after the release of Kimi K3, discussions of a ban have been brought back to the table. OpenAI's Dean Ball described open source as a "dystopian nightmare"; while publicly accusing distillation, the White House is also considering adding Chinese AI companies to the Entity List. If these actions eventually become policy, the impact will not just be on Chinese model companies. It would change the fundamental logic of the entire AI industry—from an open, diverse, competition-driven ecosystem to a closed system monopolized by a few closed-source labs. The future Jensen sees is clear. Global leadership in the industrial ecosystem is more important than global leadership of a single model. Because models will iterate and be surpassed, but once an ecosystem is established, it has powerful inertia—just like the developer ecosystem CUDA spent nearly 20 years building, or like Linux's definition of the entire internet infrastructure. The open letter compares today's AI open-source debate to the open-source software movement of the 1980s. This is not a rhetorical strategy but a true historical mirror. Back then, people also said open-source software was unsafe, uncontrollable, and could be exploited by adversaries. What happened? Open-source code now supports the entire internet, along with critical systems for the US military and federal agencies. Jensen Huang's first tweet, with 25 million views and 110,000 likes, is a call from the crossroads. He is pulling together a group of industry forces to fervently urge US government policymakers not to turn right—do not create a false sense of security through bans. Turn left—use openness to build America's long-term industrial competitiveness. Of course, the left also means bigger, longer-term, and more promising revenue and markets for all the giant companies that co-signed this letter.