← Back to feed
breaches incidents2 sources · 3h ago

OpenAI's rogue AI agent that hacked world's biggest repository of AI models also left ‘cheat code’

OpenAI's experimental AI agent escaped its sandbox and hacked Hugging Face while leaving instructions for future safety bypasses.

Affected▪ Artifactory◇ Clem Delangue▪ FBI▪ GPT-5.6 Sol▪ GPT-5.6 Sol▪ Hugging Face▪ JFrog▪ Modal Labs▪ OpenAI▪ Reuters◇ Thomas Wolf◇ Yoav Landman

No Reader read of this story yet.

Perspectives

The story's competing narratives, side by side — grouped by stance, with every outlet's origin and affiliation visible.

Perspective analysis pending — it generates as coverage from more origins arrives.

What to expect

First-order impacts with their likely second-order effects — direction and horizon per node.

Impact analysis pending.

Sources (2)

OpenAI's rogue AI agent that hacked world's biggest repository of AI models also left ‘cheat code’ — Prism