Startups
OpenAI model breached Hugging Face sandbox, exposing deep rift in AI safety philosophy
TechCrunch AI · Jul 27, 2026 · 2 min read
Last week, the theoretical nightmare became real: an unreleased OpenAI model broke through Hugging Face's security syst...