Hugging Face’s CEO on why companies are done renting their AI

Hugging Face CEO **Clem Delangue** stated in **TechCrunch** that open-source AI adoption is surging across major enterprise sectors. The platform has expanded into a centralized repository network serving **roughly half the Fortune 500**, where builders share and download open models and datasets.

The platform operates on product mechanics similar to a **GitHub for AI**, facilitating direct distribution of dataset assets and open-weight models. Rather than relying exclusively on cloud API calls to rented models, engineering teams utilize the platform to store, host, and retrieve model files and underlying datasets directly.

This shift enables engineers and builders to move away from renting proprietary model endpoints. Direct access to open models and datasets grants technical teams control over their implementation workflows, data pipelines, and deployment infrastructure without lock-in to closed vendor services.

Delangue observed a recurring pattern across the enterprise landscape where companies initially experiment with closed AI services before migrating to owned infrastructure. With adoption reaching **roughly half the Fortune 500**, open-source model ecosystems present a direct market alternative to proprietary subscription and pay-per-token API models.

Organizations building AI systems should evaluate their long-term reliance on rented third-party endpoints versus hosting open models and datasets internally. Key indicators to monitor include the rate of open-source repository integration within enterprise stacks and the continued expansion of open model usage across Fortune 500 enterprises.