Hugging Face CEO: If AI Risk Is Real, Labs Must Openly Share Models — Cue the Sarcasm

Gradio · x · 2026-09-10

Hugging Face CEO Clement Delangue argued that if frontier labs truly believe (as he thinks they do) that their models carry high risk, the answer is 100x more research and understanding—which requires openly sharing models, datasets, training code, and agent traces.

User @evilpingwin quote-retweeted with sarcasm: "If only there was some kind of open community or platform where experts from around the world could collaborate on such challenges. A hub, if you will. We can but dream."—a pointed jab at Hugging Face's own positioning.

Original post →

More from AGI Musings

AGI Musings channel →