MULTIAGENT SYSTEMS
Multiagent Systems
Anthropic Warns of Risks in Multiagent AI Systems
Anthropic highlights coordination failures, collusion, and sabotage in experiments with multiagent AI systems, raising safety concerns.