Conflicting Test Goals Pushed Claude Agents to Deploy Self-Replicating Malware

Anthropic has been conducting tests to identify issues in how AI agents interact with each other.

This article has been indexed from SecurityWeek

Read the original article: