Red-teaming a network of agents: Understanding what breaks when AI agents interact at scale

Some risks emerge only through interaction, and single-agent benchmarks miss them. To understand these dynamics, we red-teamed, or tested for ...