November 14, 2026, (Inside AI) — Nvidia's board is searching for a permanent chief executive after Jensen Huang stepped down, leaving the company's open-source AI strategy without its most prominent advocate. Interim CEO Colette Kress, the company's longtime finance chief, now oversees a portfolio that includes the Nemotron open-weight model family and a pending $12.3 billion acquisition of Hugging Face, the open-source model repository.
The leadership change arrives as Washington weighs tighter restrictions on open-weight models. Huang, 63, had argued consistently that open and closed models serve different purposes and that both are essential to the industry. His departure raises a question Nvidia's board must answer: will the company keep spending political and commercial capital on open AI without the executive who made it a personal cause?
The stakes extend beyond Nvidia. Open models account for roughly 20 per cent of tokens processed on OpenRouter, a platform that routes developer traffic across hundreds of AI models, according to research co-authored by Frank Nagle of the MIT Initiative on the Digital Economy. Closed systems from Anthropic, Google and OpenAI take the remaining 80 per cent. The gap shows that open weights remain a minority preference even as Chinese developers such as DeepSeek and Moonshot AI release capable models.
Nvidia's commercial logic does not depend on Huang's presence. Open models let businesses, governments and researchers run AI on their own infrastructure, which drives demand for GPUs, servers and data center capacity. That incentive survives any CEO transition. What may not survive is Huang's willingness to fight regulatory battles in public.
Read: Nvidia Scandal Fuels Washington Push for Tighter China Chip Export Controls
Huang Built A Record Of Advocacy
Huang's support for open models was not occasional. In October 2024, on the BG2 podcast with Brad Gerstner, he rejected the framing of open versus closed as a binary choice. He argued that open models matter most in healthcare, finance, transportation and scientific research, where organizations need systems trained for specific needs. He also drew a distinction that open models are not finished products but foundations for specialized applications.
In December 2025, during a fireside chat with John Hamre at the Center for Strategic and International Studies in Washington, Huang acknowledged that US companies led on frontier models while China had moved ahead on open-source AI. He warned that the US should not fall behind in making AI accessible.
In March, speaking with podcaster Lex Fridman, Huang explained Nvidia's decision to release its own open-weight models under the Nemotron brand. He cited industry adoption, research and innovation, and AI moving beyond language as the three reasons. "If everything is proprietary, it's hard to do research, and it's hard to innovate on top of, around, with." (Jensen Huang, former CEO, Nvidia)
Huang also used the interview to argue against the assumption that open models are inherently unsafe, a claim he repeated in July as safety concerns intensified. In an interview with Axios, he said American businesses should be allowed to use Chinese open models and rejected the idea that they pose an inherent security threat.
The commercial case and the political case have now separated. Nvidia can keep releasing Nemotron weights and funding developer tools without lobbying against restrictions in Washington. The company's next leader may calculate that quiet support for open models is safer than Huang's public advocacy, particularly with the Justice Department reviewing the Hugging Face deal on competition grounds.
Washington's Mood Has Hardened
Regulatory pressure on open models has grown through 2026. David Sacks, a top technology adviser to President Donald Trump, warned before the midterm elections that open models could become an "endangered species" in Washington because of repeated attempts to restrict or ban them. Speaking at an AI conference in San Francisco on November 3, Sacks urged the open-source community to organize politically and accused OpenAI and Anthropic of shaping rules to favor large incumbents.
Huang's exit removes a figure with the standing to challenge those proposals directly. He had argued that treating open models as automatic security risks would curtail innovation. Policymakers pushing tighter rules, especially on models developed in China, may now face less resistance from the industry's most valuable company.
Hugging Face CEO Clement Delangue said his company will keep fighting for open-source AI regardless of the leadership change at Nvidia. In a post on X, he reaffirmed the platform's commitment. Nvidia has said Hugging Face will remain an open platform and that developers will not be required to use Nvidia hardware to build or deploy models. The transaction is expected to close in the first half of 2027, subject to regulatory approval.
Nvidia's Nemotron program appears likely to continue in the near term. Reports indicate Huang's departure is unlikely to halt model development. Whether open AI remains a strategic priority will depend on the incoming CEO and a board that must weigh regulatory risk against the commercial benefits of a broad developer ecosystem.
The open-source movement now lacks its most prominent corporate champion at the moment it needs one most. Nvidia's financial interest in open models remains intact, but advocacy is not the same as investment. The company's next leader will decide whether to spend political capital defending a strategy that benefits Nvidia's hardware business but invites scrutiny from Washington.
Sofia spent years studying how machines parse human language before pivoting to cover the systems now rewriting it. She reads technical papers the way others read news and looking for what the numbers are actually saying beneath the headlines.
