Clem Delangue, CEO of Hugging Face, says adoption of open-source artificial intelligence is accelerating. Over recent years Hugging Face has evolved into something like a GitHub for AI: a place where developers can share and download models and datasets.
According to Delangue, Hugging Face’s tools are now used by roughly half of the Fortune 500. He has repeatedly observed a pattern in which companies begin development using frontier, paid APIs, but as projects scale the rising costs push them toward open-source models and self-hosted deployments.
Why this matters
The shift is driven in part by economics: usage-based API costs can grow quickly as systems scale, making it more economical for larger organizations to run open-source models themselves. Delangue also expressed concern about the possibility that a small number of large companies could end up exercising disproportionate control over AI technology and its distribution.
Discussion on TechCrunch’s Equity podcast
Delangue discussed these dynamics with Rebecca Bellan on TechCrunch’s Equity podcast. The conversation addressed the open versus closed source debate in the wake of Anthropic’s halted Fable release, covering developer and corporate motivations, the role of costs, and the potential consequences of concentrated control in the AI ecosystem.
Delangue’s view is that the growth of open-source models is not only a technological trend but also a logical response to cost structures and operational choices: when a solution becomes heavily used and expensive, organizations often opt for more flexible, locally operated alternatives.
Conclusions
The discussion highlights that the transformation of the AI ecosystem is both technological and strategic. Wider adoption of open-source models can reduce dependency on external providers but raises questions about who will ultimately control critical AI infrastructure and how that control will be exercised.
