Artificial intelligence to environmental sustainability in urban mobility: A scoping review and a conceptual framework

Artificial intelligence (AI) has been widely employed in urban mobility, but evidence on its environmental effects remains fragmented. This scoping review maps AI applications in urban mobility and proposes a conceptual framework. A total of 71 journal articles published between 2015 and 2025 were analysed. Information on context, transport modes, AI techniques, and environmental indicators was synthesised through thematic analysis and organised into causal loop diagrams. The results show a predominance of machine learning and deep learning methods. Applications focus on demand prediction, fleet and traffic optimisation, and the monitoring of emissions and air quality, yielding gains in energy efficiency and reductions in travel times. However, substantial emission reductions only emerge when there is a strong modal shift towards public transport and active modes. The evidence remains largely correlational and is concentrated in large cities in high-income countries. The review concludes that AI can be a central component of low-carbon mobility strategies when articulated with land-use policies and data governance.