<p>The daily movement of more than 1.6 billion learners between home and school is, on any conventional measure, one of the largest and most consequential public-service operations in the world; yet it has rarely been treated as such, either by scholars of education or by scholars of mobility. This article begins from that omission. Recent advances in artificial intelligence and cognitive data analytics now make it technically possible to coordinate student transportation, learning environments and pedagogical strategies as a single ecosystem of educational mobility. The promise is real—reductions in operating cost, gains in service equity, and the recovery of time for learning—but it is inseparable from a series of governance problems that the technical literature tends to bracket: surveillance of children at scale, the displacement of pedagogical values by what algorithms happen to be able to measure, and the encoding of historical inequities into routing decisions presented as neutral. Drawing on transportation studies, learning analytics, mobility-justice theory and critical data studies, the article advances a single thesis: AI-driven educational mobility can be made to serve educational justice, but only if the governance frame is treated as primary and the optimisation problem as derivative, not the other way around. The argument is illustrated through the Boston Public Schools start-time and bus-routing reform documented by Bertsimas, Delarue and Martin, and concludes with the design conditions—privacy, contestability, equity audit—under which intelligent mobility systems become an instrument of educational quality rather than a vector of its erosion.</p>