A new AI framework called BEHAVE models collective human dynamics as complex mathematical systems by analyzing physical signals like body position, velocity, and gesture rather than treating individuals in isolation. The approach detects early warning signs of group stability or breakdown by mapping behavioral fields across interaction spaces, with potential applications in negotiation, crowd safety, crisis management, and clinical settings.
Why it matters: Real-time prediction of collective breakdown—in negotiations, crowds, or teams—could prevent escalation before it happens, representing a significant shift from reactive to predictive behavioral modeling.