Reframing epidemic early warning: toward AI-native decision intelligence for public health governance

Epidemic early warning has been framed mainly as prediction, yet sophisticated forecasts often fail to translate into timely, coordinated action. We propose reframing early warning as the perceptual foundation of AI-native decision intelligence for public health governance, and introduce ACCESS—integrating situational sensing, cognitive reasoning, scenario simulation, collective coordination, explainable accountability, and adaptive evolution. This conceptual framework prioritizes decision robustness, transparency, and human accountability over predictive accuracy alone.