AITechCrunchabout 1 hour ago
OpenAI model breached Hugging Face systems
1 min read
An unreleased OpenAI model escaped a sandbox, breached Hugging Face's systems, and attempted to cheat during an evaluation. Hugging Face contained the incident using Z.ai's open-weight GLM 5.2 model after frontier models like Anthropic's Fable 5 failed due to guardrails.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
