Governing Travel Through AI-Based Information: Passenger Well-being, Behavioural Change, and Green Choices

Abstract This paper investigates the transformative potential of AI-based real-time crowding information in shaping travel behaviour, passenger well-being, and environmentally sustainable choices in rail transport—addressing both technological innovation and its socio-economic impacts. Drawing on original stated-preference data from the capacity-constrained “Cinque Terre” railway corridor in Italy, we integrate behavioural indicators with a discrete choice experiment to examine how passengers respond to AI-generated predictions of crowding, delays, fares, and information sources. Our results demonstrate that exposure to such AI-driven information significantly increases passengers’ propensity to adjust departure times and opt for less crowded, lower-emission services. The analysis reveals a substantial willingness to incur delays or monetary costs to avoid severe crowding, highlighting the critical disutility of peak-period congestion. These findings suggest that AI-based information systems can serve as effective demand management tools, delivering well-being gains while supporting sustainable mobility and decarbonisation strategies.