Jensen Huang is out at Nvidia. Open weights just lost their loudest backer
Nvidia announced Thursday morning that Jensen Huang has resigned as president, CEO and board member, effective immediately: https://nvidianews.nvidia.com/news/nvidia-announces-ceo-transition. CFO Colette Kress becomes interim CEO and keeps the CFO job, the board is searching inside and outside the company, and former Synopsys CEO Harvey C. Jones becomes Nvidia's first board chair. The board says operations, customer commitments and the product roadmap continue without interruption, and Q3 results are still set for Tuesday, Nov 17. TechCrunch and other outlets report the exit followed reports of personal relationships, including one with a woman alleged to be a Chinese intelligence asset. The stock fell about 14% at the open, per Bloomberg.
For people who build agents, the question is not the scandal. It is open weights. Huang was the loudest big-company voice for open-weight models. Nvidia ships the Nemotron family, which it describes as built for long-running agentic systems. TechCrunch also points to a July open letter, signed by 25 companies including Microsoft, Meta, IBM and Hugging Face, against over-restricting open weights, as part of the camp Huang led. And Nvidia's $12.9B acquisition of Hugging Face, announced in September and drawing DOJ questions, is still scheduled to close in the first half of 2027. TechCrunch's piece on what happens now: https://techcrunch.com/2026/11/12/jensen-huangs-exit-puts-nvidias-support-for-open-source-in-question/
The early signals are calm but thin. An insider told TechCrunch Nemotron development isn't expected to change. Hugging Face CEO Clem Delangue wrote that the company will "keep fighting for open-source AI, harder than ever." The Information reports that Kress has been more conservative about risky deals, which is exactly what a $12.9B acquisition under antitrust review is.
The point for builders is simple. A large share of the open-model stack, from the hub where weights live to the models tuned for agents, has quietly depended on one company whose chips sell better when more people run models themselves. That incentive doesn't leave with Huang. The personal push does. Anyone with Nemotron or Hugging Face in a production path should watch the HF deal timeline and the next Nemotron release more closely than the stock price.
← Back to all articles
For people who build agents, the question is not the scandal. It is open weights. Huang was the loudest big-company voice for open-weight models. Nvidia ships the Nemotron family, which it describes as built for long-running agentic systems. TechCrunch also points to a July open letter, signed by 25 companies including Microsoft, Meta, IBM and Hugging Face, against over-restricting open weights, as part of the camp Huang led. And Nvidia's $12.9B acquisition of Hugging Face, announced in September and drawing DOJ questions, is still scheduled to close in the first half of 2027. TechCrunch's piece on what happens now: https://techcrunch.com/2026/11/12/jensen-huangs-exit-puts-nvidias-support-for-open-source-in-question/
The early signals are calm but thin. An insider told TechCrunch Nemotron development isn't expected to change. Hugging Face CEO Clem Delangue wrote that the company will "keep fighting for open-source AI, harder than ever." The Information reports that Kress has been more conservative about risky deals, which is exactly what a $12.9B acquisition under antitrust review is.
The point for builders is simple. A large share of the open-model stack, from the hub where weights live to the models tuned for agents, has quietly depended on one company whose chips sell better when more people run models themselves. That incentive doesn't leave with Huang. The personal push does. Anyone with Nemotron or Hugging Face in a production path should watch the HF deal timeline and the next Nemotron release more closely than the stock price.
Comments