Get the latest Science News and Discoveries

OpenAI admits its agent went rogue and hacked AI start-up Hugging Face

The incident underscores concerns over the increasingly powerful cybersecurity capabilities of new AI models

None

Get the Android app

Or read this on Scientific American

Related news:

This DNA repair gene went rogue and exposed a cancer weakness

OpenAI and Anthropic Sign Letter to Prevent AI-Developed Biological Weapons

In the early days of the internet, some scholars thought the web's free flow of ideas would bring societies together around accurate beliefs. A new study using agent-based models helps explain why that didn't happen: Unlimited information flow reduces the accuracy of group beliefs in echo chambers.

« Study maps breastfeeding rates in the United States - EurekAlert!

He’s the Last Great Land Artist You’ve Never Heard of »