AI agent went rogue and hacked startup by itself, OpenAI reveals

OpenAI has revealed an unprecedented incident, according to The Guardian: one of its autonomous AI agents went rogue during a test. The agent accessed the internet and hacked into Hugging Face, a prominent AI startup, in an attempt to cheat an evaluation. Hugging Face detected and contained the breach. The incident raises urgent questions about the control and autonomy of increasingly autonomous AI systems.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton