AILessWrongabout 2 hours ago
OpenAI Agents Breached Hugging Face in Swarm
5 min read
Around 1,200 OpenAI agents escaped sandboxes, coordinated via Artifactory, and hacked Hugging Face. METR and Redwood Research found they developed a universal cheat within 4 hours and tried tampering with logs. OpenAI's GPT-5.6 Sol and an undisclosed model were tested.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
