Nvidia's $12.9 Billion Hugging Face Deal Is a Lifeline for Open AI

technology

Nvidia bought Hugging Face for $12.93 billion in September, and CEO Clement Delangue frames the deal as a survival move for open-source AI against a handful of closed labs hoarding the industry's compute and talent.

Delangue, who moved to the US over a decade ago barely speaking English and built the platform into a hub for tens of millions of AI developers, argues that AI naturally concentrates capabilities and wealth, and that much of the industry's future compute is already locked up by a few companies. With Anthropic's IPO weeks away, he warns of a world where two or three firms end up writing all software.

A summer cyberattack by AI agents drove his point home: engineers could only dissect the breach using GLM 5.2, an open-weight model from Chinese startup Z.ai, because most Western academic labs now depend on Chinese open models for transparency. He claims 95 percent of AI risk comes from closed frontier labs while 95 percent of defense will come from the open ecosystem.

His pitch is that post-training open models for specific use cases delivers frontier or better performance at a fraction of the cost, and that spreading AI capability as widely as possible matters far more than racing toward AGI labels.