Our research aims to enhance mobility data sharing by developing AI-driven methods for creating, representing, and analyzing mobility data enriched with semantic information, such as points of interest, transportation means, and weather conditions. We focus on three key areas: integrating diverse mobility and location data to create unified representations of urban regions, encoding user trajectories to provide context-rich yet compact data summaries, and using generative AI to synthesize realistic urban data when real data is limited. These approaches promise to improve urban planning, support sustainable development, and enable data-informed decision-making that benefits citizens and communities.