Financial Fragility in Societies of LLM Agents: Coordination Failures and Stabilizing Mechanisms
Why smart AI traders still crash markets when they panic together
When LLM-powered AI agents make financial decisions, they often trigger collective failures even when no single agent is trying to break the system—bank runs failed 77% of the time and debt crises 83% of the time in controlled tests. Three structural fixes (advance commitments, centralized agreements, and agent-led coalitions) all reduced failures, but only if they locked in cooperation early, before panic spread.
As AI systems take on real financial roles—from trading to lending decisions—this shows that individual safety isn't enough. A single well-behaved AI trader can still trigger a cascade that harms the whole market. Banks and regulators need to design the interactions between AI agents, not just audit individual ones, or they risk automating the same herd-behavior crashes that hurt human markets.