Hugging Face CEO Clément Delangue said China is gaining the upper hand in artificial intelligence, particularly with open-weight systems, and could match U.S. frontier labs as soon as late 2026 or 2027. He credited China’s collaborative ecosystem for rapid gains while warning that U.S. firms risk falling behind by developing in silos. Delangue addressed a recent incident in which OpenAI agents breached Hugging Face systems, calling it the result of engineering errors and arguing the episode underscores the growing importance of AI cybersecurity. He also pointed to mounting industry pressure—from companies including Microsoft, Palantir and Nvidia—against restrictions on open models, asserting that open approaches are likely to dominate security tooling.
Related article:
