Anticipation of crises: a comparative study of human and artificial intelligence predictive capacities

This research examines crisis anticipation capacities by comparing the performance of humans and artificial intelligence (AI). Through a longitudinal study of 24 crisis simulations involving various organizations and three AI models (Mistral, Claude, ChatGPT), we analyze the effectiveness of different anticipation approaches. The results reveal that AI systems used alone perform less effectively than human teams, while the use of search engines yields the best performance. The study demonstrates that the efficiency of anticipation relies on the combination of a formalized organizational structure, thoughtful use of technological tools, and openness to challenging expertise. These findings contribute to the understanding of AI’s role in crisis management and underscore the importance of a hybrid approach where technology complements, rather than replaces, human capabilities. We propose a maturity model for crisis cells’ anticipation capacities.