Nvidia promised to keep Hugging Face open to everyone when it announced plans to buy the AI platform for nearly $13 billion. Two months later, the CEO who made that promise is gone, and Hugging Face's future is looking a little less certain.
Following Jensen Huang's resignation on November 12, Hugging Face CEO Clément Delangue shared his reaction on X. He called it a "really sad day" before promising, "we'll keep fighting for open-source AI, harder than ever."
Huang was one of the biggest supporters of open AI models, and his enthusiasm played a major role in Nvidia's decision to acquire Hugging Face. Now, with new leadership taking over and the acquisition still awaiting completion, there's a question developers probably want answered.
Will Nvidia keep Hugging Face open to everyone, even if they're using a competitor's technology?
Why Nvidia wants Hugging Face in the first place
Hugging Face has become a popular destination for developers who want to build AI applications without creating their own models from scratch.
You can visit the platform, find an existing AI model, and use it for your project. Developers also share their work there, which gives others a starting point for building new applications.
According to Nvidia, Hugging Face has over 18 million users and hosts over 3 million AI models. More than 200,000 companies use its services, too.
That's a pretty big community for Nvidia to become part of.
When the chipmaker announced its $12.93 billion acquisition in September, Huang promised that Hugging Face would continue supporting competing AI models and hardware. Developers could still use cloud services from other providers. They wouldn't have to buy Nvidia GPUs just to run models available on Hugging Face.
Huang had also been advocating for open-weight AI, which allows developers to download a model's parameters and run it themselves. Open-weight models give developers access to the model's learned parameters, while fully open-source AI comes with additional requirements for access to code and information about how the model was developed.
Hugging Face supports both, and Huang saw value in giving developers more freedom to experiment. The question now is whether Nvidia's next CEO will be equally enthusiastic.
Why the government is asking questions
Nvidia hasn't announced any plans to change Hugging Face's open approach. Still, regulators are interested in what could happen if the acquisition goes through.
In October, Reuters reported that the US Department of Justice had started asking questions about the deal and its possible effects on competition.
Nvidia makes the GPUs that power many of today's AI models. If it owns Hugging Face, it would also control a platform where millions of developers go to find those models.
Developers who use AMD chips can find compatible models and tools on Hugging Face. But if Nvidia eventually gives its own hardware better placement or support, those competing options could become harder to find or use.
Nvidia hasn't announced any plans to do this. In fact, the company promised to continue supporting competing hardware when it announced the acquisition.
The Justice Department hasn't decided to block the deal either, but regulators want to understand how Nvidia's ownership could affect competition.
Nvidia has a reason to keep Hugging Face open
Huang might be gone, but Nvidia could still make plenty of money from an open AI ecosystem. After all, the company sells the chips that developers use to train AI models and run applications.
Every new AI application needs computing power. If more developers build their own models, Nvidia has more potential customers for its GPUs.
Open AI helps encourage that activity because developers can start with existing models instead of spending enormous amounts of money creating everything themselves. Nvidia has also released its own Nemotron family of open-weight models, so its interest in open AI goes beyond the Hugging Face deal.
There's another reason Nvidia might want to keep developers happy.
Google has developed its own AI chips, and other major technology companies are investing in alternatives to Nvidia hardware. Supporting a large community of independent AI developers could help Nvidia reach customers beyond the tech giants that are building competing chips.
As The Neuron previously explored, buying Hugging Face could also help Nvidia develop closer relationships with the people who build AI applications. Of course, Nvidia could make things uncomfortable for itself if developers start feeling pressured to use its hardware.
Some might simply move to other platforms.
And after agreeing to spend nearly $13 billion on Hugging Face, losing the community that made it valuable would be a pretty expensive problem.
So, what happens to Hugging Face now?
Nvidia has appointed longtime chief financial officer Colette Kress as interim CEO while its board searches for a permanent replacement. The company said in its November 12 announcement that customer commitments and existing product plans were expected to continue.
However, it didn't specifically mention Hugging Face or explain whether the leadership change would affect the acquisition. The deal is still expected to close in the first half of 2027, assuming it receives the necessary approvals.
For now, there's little indication that anything is changing for Hugging Face users. The company continues to operate, and Delangue has made it clear that he wants to defend open AI.
But developers will probably want more than a promise once Nvidia takes ownership.
They'll want to see whether competing AI models remain easy to find and whether Hugging Face continues supporting hardware from companies like AMD.
It'll also be worth watching how Nvidia's next CEO talks about open AI. Huang was unusually vocal about supporting it, and whoever replaces him will have to decide how much of his approach to carry forward.
Hugging Face's CEO says he's ready to keep fighting for open AI. Now developers will be watching to see whether Nvidia's next CEO is on the same page.
